diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..f7b987cf7576940593252468e52cb9eba316e258 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +train.log filter=lfs diff=lfs merge=lfs -text diff --git a/checkpoint-100000/config.json b/checkpoint-100000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d7b77e4255263f3803a4b6f94ca6831e459b70 --- /dev/null +++ b/checkpoint-100000/config.json @@ -0,0 +1,64 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/checkpoint-100000/experiment_cfg/metadata.json b/checkpoint-100000/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..300eb070963337016e16465f50ba8351ea2eb25c --- /dev/null +++ b/checkpoint-100000/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.3139495849609375, + 0.4876587688922882, + 0.7196521759033203 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6996827125549316 + ], + "mean": [ + 2.749288365724937, + -2.0455216239909983, + 0.700532551020533 + ], + "std": [ + 1.6786295897046262, + 1.312897969434244, + 0.00133751758906258 + ], + "q01": [ + -1.4241907881148037, + -5.1716547935950885, + 0.6999670898768614 + ], + "q99": [ + 6.198111545831555, + -0.5873531334104489, + 0.7038833014242587 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.23084999587450247, + 0.6238431088581847 + ], + "std": [ + 0.0, + 0.0, + 0.6556404371644047, + 0.3572252105732042 + ], + "q01": [ + 0.0, + 0.0, + -0.9999999999999999, + 1.7482354593273349e-06 + ], + "q99": [ + 0.0, + 0.0, + 0.9999999999999999, + 0.9999999999999999 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8003877401351929, + 0.9829942584037781 + ], + "min": [ + -0.37471598386764526, + -0.8472502827644348, + -0.25070279836654663 + ], + "mean": [ + 0.28667332601863704, + -0.038315473170876704, + 0.45974658906777444 + ], + "std": [ + 0.17067058828305154, + 0.23569952037784195, + 0.21950702354181487 + ], + "q01": [ + 0.015172948963784228, + -0.44450502063220615, + 0.23314444103329338 + ], + "q99": [ + 0.5609069689572412, + 0.38454179128009547, + 0.7028102736473852 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9984455108642578, + 0.9434727430343628, + 0.9062229990959167 + ], + "min": [ + -0.9999930262565613, + -0.9988337755203247, + -0.9618149995803833, + 2.1872274658107926e-07 + ], + "mean": [ + -0.25672297401336, + 0.024425335806687216, + -0.08740977164483847, + 0.16608739644018397 + ], + "std": [ + 0.7932585208392383, + 0.3102260195580416, + 0.3772472576315148, + 0.17451959300158773 + ], + "q01": [ + -0.99379006987882, + -0.5478102050038828, + -0.6101777952971636, + 0.002274085014134748 + ], + "q99": [ + 0.8804856038896288, + 0.5910611092873037, + 0.5216058042993562, + 0.5231965974012255 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011436971835792065, + -0.05664053186774254 + ], + "mean": [ + 0.031651551516052555, + -0.03161481387920421 + ], + "std": [ + 0.013119769222217526, + 0.01306078225192402 + ], + "q01": [ + 0.0060999246446777336, + -0.04062601653964198 + ], + "q99": [ + 0.04054891621517283, + -0.0059163567113456345 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.008144896753718373, + -0.00018893135146078263, + -0.0008739062727221845, + 0.0 + ], + "std": [ + 0.11030045424364411, + 0.10148082570313594, + 0.08975698373467506, + 0.0 + ], + "q01": [ + -0.07287373067581518, + -0.09446994199569948, + -0.07702818259249572, + 0.0 + ], + "q99": [ + 0.04485485414407898, + 0.09626941605965177, + 0.07610665906122742, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9221974313425939 + ], + "std": [ + 0.38670194342612674 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.384693946883099 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.01976129923017491, + -0.020870631344490034, + -0.05993476517349181 + ], + "std": [ + 0.4347024990804053, + 0.4221662976569997, + 0.3845506843044066 + ], + "q01": [ + -0.8835792939767807, + -0.9280919251713778, + -0.783361071470956 + ], + "q99": [ + 0.7622084438449307, + 0.8625125252609077, + 0.7470974753250706 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.008089673068544335, + -0.026498254627927348, + 0.0029083551743059005 + ], + "std": [ + 0.11216734503733773, + 0.13206002613798629, + 0.12615870412692864 + ], + "q01": [ + -0.2814627972846672, + -0.4342386516115439, + -0.34489755628340574 + ], + "q99": [ + 0.31879353266939386, + 0.28411797683220563, + 0.3597718671669894 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.3824228874769045 + ], + "std": [ + 0.9240015096469786 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.5597272305813592 + ] + } + } + }, + "modalities": { + "video": { + "robot0_eye_in_hand": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_left": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_right": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": true, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + }, + "end_effector_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": true, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/checkpoint-100000/model-00001-of-00002.safetensors b/checkpoint-100000/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c01920b209006553ac2e285ec041ad90ca2dec30 --- /dev/null +++ b/checkpoint-100000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6367df727fcd3c800408845100d656b9c74e278785c973a02c54dec8ed933396 +size 4999367032 diff --git a/checkpoint-100000/model-00002-of-00002.safetensors b/checkpoint-100000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6c952e69882086541416198136cd13b246b46213 --- /dev/null +++ b/checkpoint-100000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5cd11ad5da02d01f5de6ca62f8e51899b3bbcec45947efb2513a5c4386d64b9c +size 2586705312 diff --git a/checkpoint-100000/model.safetensors.index.json b/checkpoint-100000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/checkpoint-100000/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-100000/optimizer.pt b/checkpoint-100000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..8f0c693f6b822725843d92efd2c8e378d60db8fa --- /dev/null +++ b/checkpoint-100000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7c02e8510f3f4d877c3f9e1072a320fcd0b5991e090a6b9d2a93abe3a11391fa +size 8550719959 diff --git a/checkpoint-100000/rng_state_0.pth b/checkpoint-100000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..35ae682eabc407a49268f28e339e1632093aad10 --- /dev/null +++ b/checkpoint-100000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:53a482df1339a2fb95c62d545c35560ca5589ad45781e28de4d082d029ab02f0 +size 14917 diff --git a/checkpoint-100000/rng_state_1.pth b/checkpoint-100000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..3541ec9437ad98f86119d0a842332bbe1bb3308a --- /dev/null +++ b/checkpoint-100000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a1384222094f436066459827e07aff8e88a345ff9df8a5fe42212ccbe094c42 +size 14917 diff --git a/checkpoint-100000/scheduler.pt b/checkpoint-100000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..411c80eb9e866a0327e46741f71cead46dfba224 --- /dev/null +++ b/checkpoint-100000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5d3378f08f468f4f05299cb8aee1e5a28dc518a023f90c196f0f2a35074b994e +size 1465 diff --git a/checkpoint-100000/trainer_state.json b/checkpoint-100000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..7d67c0771ea1ce34376f2f8d68b126b519bfb8f0 --- /dev/null +++ b/checkpoint-100000/trainer_state.json @@ -0,0 +1,70034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 29.09514111143439, + "eval_steps": 500, + "global_step": 100000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002909514111143439, + "grad_norm": 5.5908684730529785, + "learning_rate": 4.5e-08, + "loss": 0.5668, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.084368705749512, + "learning_rate": 9.5e-08, + "loss": 0.5781, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 5.755157947540283, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5759, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.174816608428955, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5202, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 6.199656963348389, + "learning_rate": 2.45e-07, + "loss": 0.5455, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.251872539520264, + "learning_rate": 2.95e-07, + "loss": 0.5507, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 5.606583595275879, + "learning_rate": 3.45e-07, + "loss": 0.5264, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.967868328094482, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.5166, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.92909574508667, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4578, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.306433200836182, + "learning_rate": 4.95e-07, + "loss": 0.4091, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.75846529006958, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4072, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.2030856609344482, + "learning_rate": 5.95e-07, + "loss": 0.2992, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.5756216049194336, + "learning_rate": 6.45e-07, + "loss": 0.2703, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.3420923948287964, + "learning_rate": 6.95e-07, + "loss": 0.2492, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 0.9191234111785889, + "learning_rate": 7.45e-07, + "loss": 0.2302, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.662022054195404, + "learning_rate": 7.95e-07, + "loss": 0.1993, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.704409658908844, + "learning_rate": 8.45e-07, + "loss": 0.1814, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.6104884147644043, + "learning_rate": 8.95e-07, + "loss": 0.1694, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.4955693781375885, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1674, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.707667887210846, + "learning_rate": 9.95e-07, + "loss": 0.1507, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.6977017521858215, + "learning_rate": 1.045e-06, + "loss": 0.1547, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.49085867404937744, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1536, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.6363996267318726, + "learning_rate": 1.145e-06, + "loss": 0.1565, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.4633055627346039, + "learning_rate": 1.195e-06, + "loss": 0.1641, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5155240297317505, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1513, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5815547108650208, + "learning_rate": 1.295e-06, + "loss": 0.1431, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.5865835547447205, + "learning_rate": 1.345e-06, + "loss": 0.1419, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.5872474312782288, + "learning_rate": 1.395e-06, + "loss": 0.1326, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.4897519052028656, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1326, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.46316635608673096, + "learning_rate": 1.495e-06, + "loss": 0.1442, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.5359047651290894, + "learning_rate": 1.545e-06, + "loss": 0.1394, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.5805708169937134, + "learning_rate": 1.595e-06, + "loss": 0.1231, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.5488415360450745, + "learning_rate": 1.645e-06, + "loss": 0.127, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.41755908727645874, + "learning_rate": 1.695e-06, + "loss": 0.1375, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.6258692145347595, + "learning_rate": 1.745e-06, + "loss": 0.1202, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.5017617344856262, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1317, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.710078239440918, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.46977895498275757, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1215, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.4802077114582062, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.116, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.6448793411254883, + "learning_rate": 1.995e-06, + "loss": 0.1172, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4394509494304657, + "learning_rate": 2.045e-06, + "loss": 0.1145, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.6547806859016418, + "learning_rate": 2.095e-06, + "loss": 0.1174, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.6954914331436157, + "learning_rate": 2.145e-06, + "loss": 0.1245, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.670100212097168, + "learning_rate": 2.195e-06, + "loss": 0.1133, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.5230806469917297, + "learning_rate": 2.245e-06, + "loss": 0.11, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.46809589862823486, + "learning_rate": 2.295e-06, + "loss": 0.1069, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.6178835034370422, + "learning_rate": 2.345e-06, + "loss": 0.1099, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.41567298769950867, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1072, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4968737065792084, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1093, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.46893924474716187, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1046, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.5041665434837341, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1014, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5833673477172852, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1095, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.5565881133079529, + "learning_rate": 2.645e-06, + "loss": 0.1057, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6213339567184448, + "learning_rate": 2.695e-06, + "loss": 0.1087, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.5638794898986816, + "learning_rate": 2.745e-06, + "loss": 0.099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.5594047904014587, + "learning_rate": 2.795e-06, + "loss": 0.1065, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5521039366722107, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1072, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.5013270378112793, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1045, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.44355231523513794, + "learning_rate": 2.945e-06, + "loss": 0.0985, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.5467342734336853, + "learning_rate": 2.995e-06, + "loss": 0.101, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5065611600875854, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0939, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.5587425827980042, + "learning_rate": 3.095e-06, + "loss": 0.0969, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5561433434486389, + "learning_rate": 3.145e-06, + "loss": 0.1009, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.5576680302619934, + "learning_rate": 3.195e-06, + "loss": 0.1006, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.7426344752311707, + "learning_rate": 3.245e-06, + "loss": 0.0959, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.64913409948349, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.0962, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.5271337628364563, + "learning_rate": 3.345e-06, + "loss": 0.0927, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.6992810368537903, + "learning_rate": 3.395e-06, + "loss": 0.0924, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.559481680393219, + "learning_rate": 3.445e-06, + "loss": 0.0886, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7303935885429382, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0918, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.45654281973838806, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0934, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.41098082065582275, + "learning_rate": 3.595e-06, + "loss": 0.0911, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.43506675958633423, + "learning_rate": 3.645e-06, + "loss": 0.0887, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5131197571754456, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0983, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.4571124017238617, + "learning_rate": 3.745e-06, + "loss": 0.0928, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.6749465465545654, + "learning_rate": 3.795e-06, + "loss": 0.0926, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.46945369243621826, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1016, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.6481541395187378, + "learning_rate": 3.895e-06, + "loss": 0.0917, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.6769195795059204, + "learning_rate": 3.945e-06, + "loss": 0.0924, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5225585103034973, + "learning_rate": 3.995e-06, + "loss": 0.0909, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.642867922782898, + "learning_rate": 4.045e-06, + "loss": 0.0829, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.4617709219455719, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0924, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.5350925326347351, + "learning_rate": 4.145e-06, + "loss": 0.0821, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.4871961176395416, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0896, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.5545480847358704, + "learning_rate": 4.245e-06, + "loss": 0.0891, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.6057930588722229, + "learning_rate": 4.295e-06, + "loss": 0.0876, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.515282154083252, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0873, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.5402254462242126, + "learning_rate": 4.395e-06, + "loss": 0.0908, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4669134318828583, + "learning_rate": 4.445e-06, + "loss": 0.0837, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.46819189190864563, + "learning_rate": 4.495e-06, + "loss": 0.0888, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35766035318374634, + "learning_rate": 4.545e-06, + "loss": 0.0829, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.6769008040428162, + "learning_rate": 4.595e-06, + "loss": 0.0817, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5884619355201721, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0892, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5613101720809937, + "learning_rate": 4.695e-06, + "loss": 0.0852, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.47087088227272034, + "learning_rate": 4.745e-06, + "loss": 0.0818, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.4887458384037018, + "learning_rate": 4.795e-06, + "loss": 0.0858, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.44256526231765747, + "learning_rate": 4.845e-06, + "loss": 0.0833, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5375688076019287, + "learning_rate": 4.895e-06, + "loss": 0.075, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.6297019124031067, + "learning_rate": 4.945e-06, + "loss": 0.0932, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.6576125621795654, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0829, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.724773108959198, + "learning_rate": 5.045e-06, + "loss": 0.0813, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.54575115442276, + "learning_rate": 5.095e-06, + "loss": 0.0873, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.5186223387718201, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0736, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.46651381254196167, + "learning_rate": 5.195e-06, + "loss": 0.0791, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.5323564410209656, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0753, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.42579638957977295, + "learning_rate": 5.295e-06, + "loss": 0.0755, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.5063076019287109, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0767, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.4558819532394409, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0807, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.6920697093009949, + "learning_rate": 5.445e-06, + "loss": 0.079, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.4706108868122101, + "learning_rate": 5.495e-06, + "loss": 0.079, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.46705493330955505, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0745, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4643767774105072, + "learning_rate": 5.595e-06, + "loss": 0.0778, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.4800759553909302, + "learning_rate": 5.645e-06, + "loss": 0.0792, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.37705323100090027, + "learning_rate": 5.695e-06, + "loss": 0.0815, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.6651142239570618, + "learning_rate": 5.745e-06, + "loss": 0.0732, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.5275036692619324, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0758, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.45047008991241455, + "learning_rate": 5.845e-06, + "loss": 0.0808, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.6931971311569214, + "learning_rate": 5.895e-06, + "loss": 0.0808, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.7053916454315186, + "learning_rate": 5.945e-06, + "loss": 0.0747, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.6798260807991028, + "learning_rate": 5.995e-06, + "loss": 0.0829, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.5724003314971924, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0749, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.5836488008499146, + "learning_rate": 6.095e-06, + "loss": 0.0759, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.40817850828170776, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0778, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.41070470213890076, + "learning_rate": 6.195e-06, + "loss": 0.0765, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.43642961978912354, + "learning_rate": 6.245e-06, + "loss": 0.0711, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.49526578187942505, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0713, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.7773705720901489, + "learning_rate": 6.345e-06, + "loss": 0.0744, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.5485936999320984, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0732, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.5593521595001221, + "learning_rate": 6.445e-06, + "loss": 0.0747, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.3667270839214325, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0723, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.5555917620658875, + "learning_rate": 6.545e-06, + "loss": 0.0726, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.38107141852378845, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0647, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.46452224254608154, + "learning_rate": 6.645e-06, + "loss": 0.0787, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6018183827400208, + "learning_rate": 6.695e-06, + "loss": 0.0718, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.44997578859329224, + "learning_rate": 6.745e-06, + "loss": 0.0758, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45717573165893555, + "learning_rate": 6.795e-06, + "loss": 0.0738, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.5185182690620422, + "learning_rate": 6.845e-06, + "loss": 0.0768, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.5135819315910339, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.43529245257377625, + "learning_rate": 6.945e-06, + "loss": 0.0691, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.49389752745628357, + "learning_rate": 6.995e-06, + "loss": 0.0678, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.4113156199455261, + "learning_rate": 7.045e-06, + "loss": 0.0651, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.4072611927986145, + "learning_rate": 7.095e-06, + "loss": 0.0691, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.4524228870868683, + "learning_rate": 7.145e-06, + "loss": 0.0688, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.6225605607032776, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0676, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4841763377189636, + "learning_rate": 7.245e-06, + "loss": 0.0678, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.46297937631607056, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0759, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.5585324168205261, + "learning_rate": 7.345e-06, + "loss": 0.0671, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.57328200340271, + "learning_rate": 7.395e-06, + "loss": 0.0717, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.5072467923164368, + "learning_rate": 7.445000000000001e-06, + "loss": 0.071, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.4630240797996521, + "learning_rate": 7.495e-06, + "loss": 0.0674, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.35908272862434387, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0704, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.5117342472076416, + "learning_rate": 7.595e-06, + "loss": 0.0675, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.5214630961418152, + "learning_rate": 7.645e-06, + "loss": 0.0702, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.39036089181900024, + "learning_rate": 7.695e-06, + "loss": 0.0709, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.558722734451294, + "learning_rate": 7.745e-06, + "loss": 0.0705, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.3612176179885864, + "learning_rate": 7.795e-06, + "loss": 0.0715, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.523444652557373, + "learning_rate": 7.845e-06, + "loss": 0.0685, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.5231291651725769, + "learning_rate": 7.895e-06, + "loss": 0.0717, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.50368332862854, + "learning_rate": 7.945e-06, + "loss": 0.0694, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.4126806855201721, + "learning_rate": 7.995e-06, + "loss": 0.0693, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.719352662563324, + "learning_rate": 8.045e-06, + "loss": 0.0651, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.4405677020549774, + "learning_rate": 8.095e-06, + "loss": 0.0645, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.34746286273002625, + "learning_rate": 8.145e-06, + "loss": 0.0687, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.43200597167015076, + "learning_rate": 8.195e-06, + "loss": 0.0655, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.3464098274707794, + "learning_rate": 8.245e-06, + "loss": 0.0639, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.27263233065605164, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0629, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.343181848526001, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3949901759624481, + "learning_rate": 8.395e-06, + "loss": 0.0643, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.4489574730396271, + "learning_rate": 8.445e-06, + "loss": 0.0698, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.4746992588043213, + "learning_rate": 8.495e-06, + "loss": 0.0702, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.3651655316352844, + "learning_rate": 8.545e-06, + "loss": 0.0673, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.471071720123291, + "learning_rate": 8.595e-06, + "loss": 0.0648, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.35623008012771606, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0698, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.41837432980537415, + "learning_rate": 8.695e-06, + "loss": 0.0639, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.45608940720558167, + "learning_rate": 8.745e-06, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.40346869826316833, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0625, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.479824960231781, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0703, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.38570791482925415, + "learning_rate": 8.895e-06, + "loss": 0.0643, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.36530473828315735, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0652, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37868866324424744, + "learning_rate": 8.995000000000001e-06, + "loss": 0.068, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.32546791434288025, + "learning_rate": 9.045e-06, + "loss": 0.0636, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4664970338344574, + "learning_rate": 9.095e-06, + "loss": 0.0671, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4767569601535797, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0625, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.41823142766952515, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0685, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.3709985613822937, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0681, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.3933296799659729, + "learning_rate": 9.295e-06, + "loss": 0.0678, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.3079732060432434, + "learning_rate": 9.345e-06, + "loss": 0.0598, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.33485978841781616, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0644, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.40312913060188293, + "learning_rate": 9.445e-06, + "loss": 0.0684, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.4334392547607422, + "learning_rate": 9.495e-06, + "loss": 0.0622, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.5787541270256042, + "learning_rate": 9.545e-06, + "loss": 0.0656, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.2827294170856476, + "learning_rate": 9.595e-06, + "loss": 0.0628, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4431678056716919, + "learning_rate": 9.645e-06, + "loss": 0.0633, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.3916102647781372, + "learning_rate": 9.695e-06, + "loss": 0.0632, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.41470348834991455, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0612, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.3224067986011505, + "learning_rate": 9.795e-06, + "loss": 0.061, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.4110186696052551, + "learning_rate": 9.845e-06, + "loss": 0.066, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.4786360561847687, + "learning_rate": 9.895e-06, + "loss": 0.0648, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3667179048061371, + "learning_rate": 9.945e-06, + "loss": 0.0663, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.31722185015678406, + "learning_rate": 9.995e-06, + "loss": 0.0588, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.29271745681762695, + "learning_rate": 1.0045e-05, + "loss": 0.0636, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.3536592423915863, + "learning_rate": 1.0095e-05, + "loss": 0.0625, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.3228510916233063, + "learning_rate": 1.0145e-05, + "loss": 0.067, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.38474392890930176, + "learning_rate": 1.0195e-05, + "loss": 0.0609, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.40410760045051575, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0638, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.3632320463657379, + "learning_rate": 1.0295e-05, + "loss": 0.0587, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.37775692343711853, + "learning_rate": 1.0345e-05, + "loss": 0.0673, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.3343006372451782, + "learning_rate": 1.0395e-05, + "loss": 0.0599, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3232511281967163, + "learning_rate": 1.0445e-05, + "loss": 0.0661, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.38730430603027344, + "learning_rate": 1.0495e-05, + "loss": 0.0649, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.38972100615501404, + "learning_rate": 1.0545e-05, + "loss": 0.0582, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.27242428064346313, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0583, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.30514073371887207, + "learning_rate": 1.0645e-05, + "loss": 0.065, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.45364341139793396, + "learning_rate": 1.0695e-05, + "loss": 0.0576, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.3863564729690552, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0579, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.33717092871665955, + "learning_rate": 1.0795e-05, + "loss": 0.0565, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38073790073394775, + "learning_rate": 1.0845e-05, + "loss": 0.0585, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4277617037296295, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0622, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.3545698821544647, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0609, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.30646640062332153, + "learning_rate": 1.0995e-05, + "loss": 0.0659, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.3607352375984192, + "learning_rate": 1.1045e-05, + "loss": 0.0621, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3624425530433655, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0626, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3856692612171173, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0604, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.4390200972557068, + "learning_rate": 1.1195e-05, + "loss": 0.0601, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.4108872711658478, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0618, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.32732832431793213, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0636, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.3602239787578583, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.29978248476982117, + "learning_rate": 1.1395e-05, + "loss": 0.058, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.48490920662879944, + "learning_rate": 1.1445e-05, + "loss": 0.0614, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.3278113901615143, + "learning_rate": 1.1495e-05, + "loss": 0.0611, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.37862712144851685, + "learning_rate": 1.1545e-05, + "loss": 0.0626, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.3514714241027832, + "learning_rate": 1.1595e-05, + "loss": 0.0583, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.27256545424461365, + "learning_rate": 1.1645e-05, + "loss": 0.062, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.3305833339691162, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0572, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.2881031930446625, + "learning_rate": 1.1745e-05, + "loss": 0.0641, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.3219377398490906, + "learning_rate": 1.1795e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.34189504384994507, + "learning_rate": 1.1845e-05, + "loss": 0.0607, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30541297793388367, + "learning_rate": 1.1895e-05, + "loss": 0.056, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.35898879170417786, + "learning_rate": 1.1945e-05, + "loss": 0.0598, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.33221277594566345, + "learning_rate": 1.1995e-05, + "loss": 0.0603, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.3435590863227844, + "learning_rate": 1.2045e-05, + "loss": 0.0553, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.24820426106452942, + "learning_rate": 1.2095e-05, + "loss": 0.0644, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.3749466836452484, + "learning_rate": 1.2145e-05, + "loss": 0.0615, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.3984697759151459, + "learning_rate": 1.2195e-05, + "loss": 0.0595, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3250048756599426, + "learning_rate": 1.2245e-05, + "loss": 0.0608, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.2794302701950073, + "learning_rate": 1.2295e-05, + "loss": 0.0578, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3811834752559662, + "learning_rate": 1.2345e-05, + "loss": 0.0612, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3160156309604645, + "learning_rate": 1.2395e-05, + "loss": 0.0541, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.3825189173221588, + "learning_rate": 1.2445e-05, + "loss": 0.0621, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.3966415226459503, + "learning_rate": 1.2495e-05, + "loss": 0.0598, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3615199625492096, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0616, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.440762996673584, + "learning_rate": 1.2595e-05, + "loss": 0.0599, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.345599889755249, + "learning_rate": 1.2645e-05, + "loss": 0.0597, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.2963278889656067, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0556, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.3720734417438507, + "learning_rate": 1.2745e-05, + "loss": 0.0618, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3937445878982544, + "learning_rate": 1.2795e-05, + "loss": 0.0624, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.4492146372795105, + "learning_rate": 1.2845e-05, + "loss": 0.0573, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.5172158479690552, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0594, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.3796320855617523, + "learning_rate": 1.2945e-05, + "loss": 0.0528, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.42423033714294434, + "learning_rate": 1.2995e-05, + "loss": 0.0615, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.4578545093536377, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0562, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.38345491886138916, + "learning_rate": 1.3095e-05, + "loss": 0.0574, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.2539816200733185, + "learning_rate": 1.3145e-05, + "loss": 0.0562, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.44323891401290894, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0588, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.4074377417564392, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0591, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.5252887606620789, + "learning_rate": 1.3295e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.39536044001579285, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0588, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.6300645470619202, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0584, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.312466561794281, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0588, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3810383379459381, + "learning_rate": 1.3495e-05, + "loss": 0.0597, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.30645400285720825, + "learning_rate": 1.3545e-05, + "loss": 0.0589, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.2857605218887329, + "learning_rate": 1.3595e-05, + "loss": 0.0589, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.3206622898578644, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0597, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3226073086261749, + "learning_rate": 1.3695e-05, + "loss": 0.0559, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.24912619590759277, + "learning_rate": 1.3745e-05, + "loss": 0.0547, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.29700762033462524, + "learning_rate": 1.3795e-05, + "loss": 0.0563, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.32627588510513306, + "learning_rate": 1.3845e-05, + "loss": 0.0602, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.48569780588150024, + "learning_rate": 1.3895e-05, + "loss": 0.0553, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.323607474565506, + "learning_rate": 1.3945e-05, + "loss": 0.0578, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.42626142501831055, + "learning_rate": 1.3995e-05, + "loss": 0.0539, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.3281888961791992, + "learning_rate": 1.4045e-05, + "loss": 0.0638, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.38216397166252136, + "learning_rate": 1.4095e-05, + "loss": 0.0572, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.2816051244735718, + "learning_rate": 1.4145e-05, + "loss": 0.0543, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.2604160010814667, + "learning_rate": 1.4195e-05, + "loss": 0.0555, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.34318771958351135, + "learning_rate": 1.4245e-05, + "loss": 0.0565, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.31248190999031067, + "learning_rate": 1.4295e-05, + "loss": 0.0583, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.3014426827430725, + "learning_rate": 1.4345e-05, + "loss": 0.0586, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3650360703468323, + "learning_rate": 1.4395e-05, + "loss": 0.0581, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.33519116044044495, + "learning_rate": 1.4445e-05, + "loss": 0.0611, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.3595008850097656, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.054, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.32799217104911804, + "learning_rate": 1.4545e-05, + "loss": 0.0574, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.29410457611083984, + "learning_rate": 1.4595e-05, + "loss": 0.0582, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.2727576792240143, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0588, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.36562401056289673, + "learning_rate": 1.4695e-05, + "loss": 0.0597, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.3351168930530548, + "learning_rate": 1.4745e-05, + "loss": 0.0558, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.23880337178707123, + "learning_rate": 1.4795e-05, + "loss": 0.0571, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.3910059928894043, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0573, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.3103587031364441, + "learning_rate": 1.4895e-05, + "loss": 0.0578, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.3265620768070221, + "learning_rate": 1.4945e-05, + "loss": 0.0542, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.3060731291770935, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0537, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.41589322686195374, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0546, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.3102424740791321, + "learning_rate": 1.5095e-05, + "loss": 0.0564, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.32286810874938965, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0519, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.3129369020462036, + "learning_rate": 1.5195e-05, + "loss": 0.0566, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.43735018372535706, + "learning_rate": 1.5245e-05, + "loss": 0.0552, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.440449595451355, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0589, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.3631616532802582, + "learning_rate": 1.5345e-05, + "loss": 0.0511, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.29124221205711365, + "learning_rate": 1.5395e-05, + "loss": 0.055, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.4247622489929199, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0562, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.36976534128189087, + "learning_rate": 1.5495e-05, + "loss": 0.0604, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.3580514192581177, + "learning_rate": 1.5545e-05, + "loss": 0.0547, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.2969493567943573, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0564, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28583595156669617, + "learning_rate": 1.5645e-05, + "loss": 0.0586, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.3220328390598297, + "learning_rate": 1.5695e-05, + "loss": 0.0588, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.32270652055740356, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0559, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.3170712888240814, + "learning_rate": 1.5795e-05, + "loss": 0.0566, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.2966325879096985, + "learning_rate": 1.5845e-05, + "loss": 0.0551, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.2788766622543335, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.051, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.3521708548069, + "learning_rate": 1.5945e-05, + "loss": 0.0558, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.35105040669441223, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0551, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.4221852123737335, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0573, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.4217511713504791, + "learning_rate": 1.6095e-05, + "loss": 0.0532, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.4519175589084625, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0489, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3427059054374695, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0544, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29204127192497253, + "learning_rate": 1.6245e-05, + "loss": 0.0525, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.24138635396957397, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0559, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.3599991500377655, + "learning_rate": 1.6345e-05, + "loss": 0.0544, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.3326258361339569, + "learning_rate": 1.6395e-05, + "loss": 0.0593, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31976261734962463, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0624, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.26379555463790894, + "learning_rate": 1.6495e-05, + "loss": 0.0529, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2576923966407776, + "learning_rate": 1.6545e-05, + "loss": 0.0527, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.30890268087387085, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0571, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.4547998309135437, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.3050350248813629, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0552, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.2984176278114319, + "learning_rate": 1.6745e-05, + "loss": 0.0528, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.34583792090415955, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0512, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.33085328340530396, + "learning_rate": 1.6845e-05, + "loss": 0.0563, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.33391326665878296, + "learning_rate": 1.6895e-05, + "loss": 0.0527, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.23657914996147156, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0509, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.3719659149646759, + "learning_rate": 1.6995e-05, + "loss": 0.0568, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.4268389940261841, + "learning_rate": 1.7045e-05, + "loss": 0.0525, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.2726958394050598, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0482, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27847009897232056, + "learning_rate": 1.7145e-05, + "loss": 0.0559, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.7195e-05, + "loss": 0.0588, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.2666028141975403, + "learning_rate": 1.7245e-05, + "loss": 0.0548, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.26048436760902405, + "learning_rate": 1.7295e-05, + "loss": 0.0592, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.22569318115711212, + "learning_rate": 1.7345e-05, + "loss": 0.0505, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.35350096225738525, + "learning_rate": 1.7395e-05, + "loss": 0.0566, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.3166314363479614, + "learning_rate": 1.7445e-05, + "loss": 0.0517, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.24918541312217712, + "learning_rate": 1.7495e-05, + "loss": 0.0545, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3416207730770111, + "learning_rate": 1.7545e-05, + "loss": 0.0544, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.22633257508277893, + "learning_rate": 1.7595e-05, + "loss": 0.0545, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.3972736597061157, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0591, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3213656544685364, + "learning_rate": 1.7695e-05, + "loss": 0.0537, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.30361926555633545, + "learning_rate": 1.7745e-05, + "loss": 0.0529, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.35521000623703003, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0549, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.4444983899593353, + "learning_rate": 1.7845e-05, + "loss": 0.0529, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.32495826482772827, + "learning_rate": 1.7895e-05, + "loss": 0.0557, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.2893896698951721, + "learning_rate": 1.7945e-05, + "loss": 0.056, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.2607886493206024, + "learning_rate": 1.7995e-05, + "loss": 0.055, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.3305937647819519, + "learning_rate": 1.8045e-05, + "loss": 0.0529, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.3003319799900055, + "learning_rate": 1.8095e-05, + "loss": 0.0515, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3523712456226349, + "learning_rate": 1.8145e-05, + "loss": 0.0524, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.27575406432151794, + "learning_rate": 1.8195e-05, + "loss": 0.0535, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.36445286870002747, + "learning_rate": 1.8245e-05, + "loss": 0.0552, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.41415491700172424, + "learning_rate": 1.8295e-05, + "loss": 0.0537, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.3738953471183777, + "learning_rate": 1.8345e-05, + "loss": 0.0555, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.3446105420589447, + "learning_rate": 1.8395e-05, + "loss": 0.0526, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.39016082882881165, + "learning_rate": 1.8445e-05, + "loss": 0.0536, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.4075721502304077, + "learning_rate": 1.8495e-05, + "loss": 0.0548, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.4071345925331116, + "learning_rate": 1.8545e-05, + "loss": 0.0539, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.34446755051612854, + "learning_rate": 1.8595e-05, + "loss": 0.057, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.3431609272956848, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0515, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.31537994742393494, + "learning_rate": 1.8695e-05, + "loss": 0.0534, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.4846970736980438, + "learning_rate": 1.8745e-05, + "loss": 0.0532, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.3109683394432068, + "learning_rate": 1.8795e-05, + "loss": 0.0547, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.28903284668922424, + "learning_rate": 1.8845e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.26592350006103516, + "learning_rate": 1.8895e-05, + "loss": 0.0499, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.27347302436828613, + "learning_rate": 1.8945e-05, + "loss": 0.052, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.3194405138492584, + "learning_rate": 1.8995e-05, + "loss": 0.0521, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.40457430481910706, + "learning_rate": 1.9045e-05, + "loss": 0.0523, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.3055919408798218, + "learning_rate": 1.9095e-05, + "loss": 0.0504, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.4911873936653137, + "learning_rate": 1.9145e-05, + "loss": 0.0531, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.30619072914123535, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0525, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.24404728412628174, + "learning_rate": 1.9245e-05, + "loss": 0.0498, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.31723177433013916, + "learning_rate": 1.9295e-05, + "loss": 0.0519, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.28105342388153076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0528, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.32965341210365295, + "learning_rate": 1.9395e-05, + "loss": 0.0504, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.2487417608499527, + "learning_rate": 1.9445e-05, + "loss": 0.0552, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3788810670375824, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0535, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.2561226189136505, + "learning_rate": 1.9545e-05, + "loss": 0.056, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.29964715242385864, + "learning_rate": 1.9595e-05, + "loss": 0.0565, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.386600136756897, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0468, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.37690475583076477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.35367539525032043, + "learning_rate": 1.9745e-05, + "loss": 0.0501, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.30189844965934753, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0524, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.27525416016578674, + "learning_rate": 1.9845e-05, + "loss": 0.0524, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.22557462751865387, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.3572654128074646, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0536, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.37130483984947205, + "learning_rate": 1.9995e-05, + "loss": 0.0507, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.40820422768592834, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0547, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.24222403764724731, + "learning_rate": 2.0095e-05, + "loss": 0.0536, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.35473307967185974, + "learning_rate": 2.0145e-05, + "loss": 0.0503, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24238434433937073, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0511, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.32721346616744995, + "learning_rate": 2.0245e-05, + "loss": 0.0474, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.3787407875061035, + "learning_rate": 2.0295e-05, + "loss": 0.0515, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.3435424566268921, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0521, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.250515341758728, + "learning_rate": 2.0395e-05, + "loss": 0.0539, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.2730806767940521, + "learning_rate": 2.0445e-05, + "loss": 0.0544, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.32704228162765503, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0516, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.26865822076797485, + "learning_rate": 2.0545e-05, + "loss": 0.0557, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.31514087319374084, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.051, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.2616795301437378, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0483, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2979176938533783, + "learning_rate": 2.0695e-05, + "loss": 0.056, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.2876877188682556, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0496, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.2801620662212372, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0533, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.3521801233291626, + "learning_rate": 2.0845e-05, + "loss": 0.051, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.3016152083873749, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0515, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.29837727546691895, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0483, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.255982369184494, + "learning_rate": 2.0995e-05, + "loss": 0.0482, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.20852255821228027, + "learning_rate": 2.1045e-05, + "loss": 0.05, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.2746782898902893, + "learning_rate": 2.1095e-05, + "loss": 0.0513, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.28745001554489136, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0486, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.3163246810436249, + "learning_rate": 2.1195e-05, + "loss": 0.0536, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.2279432713985443, + "learning_rate": 2.1245e-05, + "loss": 0.0497, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.2273164689540863, + "learning_rate": 2.1295e-05, + "loss": 0.0485, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.3124007284641266, + "learning_rate": 2.1345e-05, + "loss": 0.0473, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.2097717970609665, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0514, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.24428582191467285, + "learning_rate": 2.1445e-05, + "loss": 0.0543, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.47236987948417664, + "learning_rate": 2.1495e-05, + "loss": 0.0525, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.3367164433002472, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0525, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.34086376428604126, + "learning_rate": 2.1595e-05, + "loss": 0.049, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.3060646951198578, + "learning_rate": 2.1645e-05, + "loss": 0.0538, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.2997030019760132, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0553, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.2652549743652344, + "learning_rate": 2.1745e-05, + "loss": 0.051, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.3072361350059509, + "learning_rate": 2.1795e-05, + "loss": 0.0516, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2494215965270996, + "learning_rate": 2.1845e-05, + "loss": 0.0476, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.20796196162700653, + "learning_rate": 2.1895e-05, + "loss": 0.0553, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.24235095083713531, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.21033242344856262, + "learning_rate": 2.1995e-05, + "loss": 0.0542, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.26056164503097534, + "learning_rate": 2.2045e-05, + "loss": 0.0546, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.24798662960529327, + "learning_rate": 2.2095e-05, + "loss": 0.0529, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.22179707884788513, + "learning_rate": 2.2145e-05, + "loss": 0.0542, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.2660025358200073, + "learning_rate": 2.2195e-05, + "loss": 0.0496, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.3771500289440155, + "learning_rate": 2.2245e-05, + "loss": 0.0502, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.3059975206851959, + "learning_rate": 2.2295e-05, + "loss": 0.0482, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.2941029667854309, + "learning_rate": 2.2345e-05, + "loss": 0.0483, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3481833040714264, + "learning_rate": 2.2395e-05, + "loss": 0.0541, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.2729704678058624, + "learning_rate": 2.2445e-05, + "loss": 0.0514, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.44202038645744324, + "learning_rate": 2.2495e-05, + "loss": 0.0534, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.311088889837265, + "learning_rate": 2.2545e-05, + "loss": 0.0474, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30663055181503296, + "learning_rate": 2.2595e-05, + "loss": 0.0491, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.3530942499637604, + "learning_rate": 2.2645e-05, + "loss": 0.0503, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.27998727560043335, + "learning_rate": 2.2695e-05, + "loss": 0.0493, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.2679221034049988, + "learning_rate": 2.2745e-05, + "loss": 0.0512, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.348821222782135, + "learning_rate": 2.2795e-05, + "loss": 0.0489, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.260958194732666, + "learning_rate": 2.2845e-05, + "loss": 0.0474, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.2429468184709549, + "learning_rate": 2.2895e-05, + "loss": 0.053, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.2635575234889984, + "learning_rate": 2.2945e-05, + "loss": 0.053, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.3416514992713928, + "learning_rate": 2.2995e-05, + "loss": 0.0476, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.221172496676445, + "learning_rate": 2.3045e-05, + "loss": 0.0519, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.30541878938674927, + "learning_rate": 2.3095e-05, + "loss": 0.0522, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.25274574756622314, + "learning_rate": 2.3145e-05, + "loss": 0.0499, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.366253525018692, + "learning_rate": 2.3195e-05, + "loss": 0.0463, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.16825024783611298, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0493, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.3158467710018158, + "learning_rate": 2.3295e-05, + "loss": 0.0482, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.2887287139892578, + "learning_rate": 2.3345e-05, + "loss": 0.053, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.4717058539390564, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0515, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.3148113489151001, + "learning_rate": 2.3445e-05, + "loss": 0.0523, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.23815515637397766, + "learning_rate": 2.3495e-05, + "loss": 0.0523, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23897255957126617, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0484, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.2680213749408722, + "learning_rate": 2.3595e-05, + "loss": 0.0485, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.29999756813049316, + "learning_rate": 2.3645e-05, + "loss": 0.0506, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.30607151985168457, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.05, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.26077136397361755, + "learning_rate": 2.3745e-05, + "loss": 0.0558, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2638765275478363, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.049, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.24569877982139587, + "learning_rate": 2.3845e-05, + "loss": 0.053, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.29622989892959595, + "learning_rate": 2.3895e-05, + "loss": 0.0517, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.2598831355571747, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.05, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.25716227293014526, + "learning_rate": 2.3995e-05, + "loss": 0.0453, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.37305381894111633, + "learning_rate": 2.4045e-05, + "loss": 0.0539, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.3151630461215973, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0516, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.296941339969635, + "learning_rate": 2.4145e-05, + "loss": 0.0509, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.31995704770088196, + "learning_rate": 2.4195e-05, + "loss": 0.0498, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2426208257675171, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0475, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.23927678167819977, + "learning_rate": 2.4295e-05, + "loss": 0.05, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.1887904703617096, + "learning_rate": 2.4345e-05, + "loss": 0.05, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.18480589985847473, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0496, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.24515867233276367, + "learning_rate": 2.4445e-05, + "loss": 0.0464, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.25505101680755615, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0481, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.24435125291347504, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0502, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.25899314880371094, + "learning_rate": 2.4595e-05, + "loss": 0.0535, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.30518922209739685, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0481, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.26098036766052246, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0477, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.24723340570926666, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.26713746786117554, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0506, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.21166373789310455, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0502, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2602159082889557, + "learning_rate": 2.4895e-05, + "loss": 0.0506, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27918240427970886, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.049, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.2131885290145874, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0477, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.28778040409088135, + "learning_rate": 2.5045e-05, + "loss": 0.0487, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.3056010603904724, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.049, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.2850024700164795, + "learning_rate": 2.5145e-05, + "loss": 0.0475, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.20490600168704987, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0497, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.32371339201927185, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0491, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.32466208934783936, + "learning_rate": 2.5295e-05, + "loss": 0.0485, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.2779853343963623, + "learning_rate": 2.5345e-05, + "loss": 0.0485, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.3257012367248535, + "learning_rate": 2.5395e-05, + "loss": 0.0486, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2625649571418762, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0486, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.19846542179584503, + "learning_rate": 2.5495e-05, + "loss": 0.0494, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.2415613830089569, + "learning_rate": 2.5545e-05, + "loss": 0.0539, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.20438073575496674, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0476, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.2543872892856598, + "learning_rate": 2.5645e-05, + "loss": 0.0489, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.26969635486602783, + "learning_rate": 2.5695e-05, + "loss": 0.0557, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.27818888425827026, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0517, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.28826993703842163, + "learning_rate": 2.5795e-05, + "loss": 0.0499, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.19518250226974487, + "learning_rate": 2.5845e-05, + "loss": 0.0499, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.2639068067073822, + "learning_rate": 2.5895e-05, + "loss": 0.0459, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.26014742255210876, + "learning_rate": 2.5945e-05, + "loss": 0.0522, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.26659756898880005, + "learning_rate": 2.5995e-05, + "loss": 0.0471, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.24073593318462372, + "learning_rate": 2.6045e-05, + "loss": 0.0488, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.2724616527557373, + "learning_rate": 2.6095e-05, + "loss": 0.0493, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.3012109398841858, + "learning_rate": 2.6145e-05, + "loss": 0.049, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23306235671043396, + "learning_rate": 2.6195e-05, + "loss": 0.0474, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.23879121243953705, + "learning_rate": 2.6245e-05, + "loss": 0.048, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20701922476291656, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0527, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.28734415769577026, + "learning_rate": 2.6345e-05, + "loss": 0.0469, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.26227426528930664, + "learning_rate": 2.6395e-05, + "loss": 0.0472, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.24196277558803558, + "learning_rate": 2.6445e-05, + "loss": 0.0462, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.3405964970588684, + "learning_rate": 2.6495e-05, + "loss": 0.0479, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.24121852219104767, + "learning_rate": 2.6545e-05, + "loss": 0.0444, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.284856379032135, + "learning_rate": 2.6595e-05, + "loss": 0.0492, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.2515311539173126, + "learning_rate": 2.6645e-05, + "loss": 0.0453, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.29297393560409546, + "learning_rate": 2.6695e-05, + "loss": 0.049, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.2528100311756134, + "learning_rate": 2.6745e-05, + "loss": 0.0477, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.29126009345054626, + "learning_rate": 2.6795e-05, + "loss": 0.0484, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.26489511132240295, + "learning_rate": 2.6845e-05, + "loss": 0.0476, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.33150023221969604, + "learning_rate": 2.6895e-05, + "loss": 0.0475, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.25904157757759094, + "learning_rate": 2.6945e-05, + "loss": 0.0495, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.24830900132656097, + "learning_rate": 2.6995e-05, + "loss": 0.0476, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.277997225522995, + "learning_rate": 2.7045e-05, + "loss": 0.0478, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.3100328743457794, + "learning_rate": 2.7095e-05, + "loss": 0.0501, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.2957077622413635, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0491, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.19402435421943665, + "learning_rate": 2.7195e-05, + "loss": 0.0426, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.20080482959747314, + "learning_rate": 2.7245e-05, + "loss": 0.0464, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.21148671209812164, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0457, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.2299162596464157, + "learning_rate": 2.7345e-05, + "loss": 0.0467, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2296467125415802, + "learning_rate": 2.7395e-05, + "loss": 0.0489, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.25099772214889526, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0487, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2503034472465515, + "learning_rate": 2.7495e-05, + "loss": 0.0533, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.24678738415241241, + "learning_rate": 2.7545e-05, + "loss": 0.0462, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.3281477987766266, + "learning_rate": 2.7595e-05, + "loss": 0.0492, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.32214412093162537, + "learning_rate": 2.7645e-05, + "loss": 0.0459, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.29244086146354675, + "learning_rate": 2.7695e-05, + "loss": 0.0495, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.2372455894947052, + "learning_rate": 2.7745e-05, + "loss": 0.0477, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.28411489725112915, + "learning_rate": 2.7795e-05, + "loss": 0.0514, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.36217001080513, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2913915514945984, + "learning_rate": 2.7895e-05, + "loss": 0.0535, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.3237503170967102, + "learning_rate": 2.7945e-05, + "loss": 0.0509, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.32016751170158386, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0526, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.25484171509742737, + "learning_rate": 2.8045e-05, + "loss": 0.0508, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.22641469538211823, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.23269280791282654, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0484, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23346202075481415, + "learning_rate": 2.8195e-05, + "loss": 0.0472, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.25522616505622864, + "learning_rate": 2.8245e-05, + "loss": 0.0493, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.2478165179491043, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0468, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.25732970237731934, + "learning_rate": 2.8345e-05, + "loss": 0.0484, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.21047812700271606, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0455, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.252676397562027, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0496, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20048962533473969, + "learning_rate": 2.8495e-05, + "loss": 0.0434, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.1821785420179367, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0502, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.24106337130069733, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0471, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.413974404335022, + "learning_rate": 2.8645e-05, + "loss": 0.0542, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.29419073462486267, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0499, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.20694036781787872, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0453, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2390667200088501, + "learning_rate": 2.8795e-05, + "loss": 0.045, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2563699781894684, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.046, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.22678422927856445, + "learning_rate": 2.8895e-05, + "loss": 0.0456, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.26104822754859924, + "learning_rate": 2.8945e-05, + "loss": 0.0446, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.22175569832324982, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0478, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.20990405976772308, + "learning_rate": 2.9045e-05, + "loss": 0.0459, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.27277135848999023, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0477, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.22422750294208527, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.047, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.2740926742553711, + "learning_rate": 2.9195e-05, + "loss": 0.0444, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2977927327156067, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0477, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26431694626808167, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0488, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.22759316861629486, + "learning_rate": 2.9345e-05, + "loss": 0.0465, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.24671563506126404, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0501, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.24065586924552917, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0439, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.3334764838218689, + "learning_rate": 2.9495e-05, + "loss": 0.0475, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.23565605282783508, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.046, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.26102808117866516, + "learning_rate": 2.9595e-05, + "loss": 0.0494, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.34544840455055237, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.05, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.2476256936788559, + "learning_rate": 2.9695e-05, + "loss": 0.0477, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.2795628309249878, + "learning_rate": 2.9745e-05, + "loss": 0.0508, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.2756715714931488, + "learning_rate": 2.9795e-05, + "loss": 0.0484, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.20578329265117645, + "learning_rate": 2.9845e-05, + "loss": 0.0461, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.20354008674621582, + "learning_rate": 2.9895e-05, + "loss": 0.0431, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24710936844348907, + "learning_rate": 2.9945e-05, + "loss": 0.0459, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.32639119029045105, + "learning_rate": 2.9995e-05, + "loss": 0.045, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.27762553095817566, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0476, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.2903423607349396, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.3292105793952942, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0503, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.23503656685352325, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0446, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.24153585731983185, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.048, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.24157844483852386, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.046, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.23378467559814453, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0478, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.20753100514411926, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0527, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.2574090361595154, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0468, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.23341813683509827, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0445, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21378666162490845, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0487, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.2410328984260559, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0488, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.22005778551101685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0478, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.3197212219238281, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.046, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.253745436668396, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0437, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.1991635262966156, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0474, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.20773734152317047, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0439, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.3026197552680969, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0462, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.22027666866779327, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0443, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.2512618899345398, + "learning_rate": 2.99997744429684e-05, + "loss": 0.044, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.23001381754875183, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0479, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.20057332515716553, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0433, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.2831740975379944, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0462, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.18362601101398468, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0472, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.29741403460502625, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0443, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.20039376616477966, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0457, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.14999084174633026, + "learning_rate": 2.999958785135414e-05, + "loss": 0.043, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.18134453892707825, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0416, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.25010257959365845, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0494, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.31700536608695984, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0455, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.25080761313438416, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0436, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.2008715718984604, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0459, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.21441133320331573, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0492, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.23232029378414154, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0459, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.30173608660697937, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0505, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.21179363131523132, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0438, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.21905283629894257, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0439, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.17932383716106415, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0452, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.21132917702198029, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0454, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.22732721269130707, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.047, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.28214913606643677, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0446, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.23979637026786804, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0454, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.3095552921295166, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0449, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.2585015594959259, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0473, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.18226519227027893, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0472, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2568792402744293, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0443, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.27914026379585266, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0423, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.21452564001083374, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0457, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.22575126588344574, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0458, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.24053922295570374, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0453, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.21956735849380493, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.048, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2325354665517807, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0439, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.1692194789648056, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.044, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.25796884298324585, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0416, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.2719244956970215, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0445, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.23312833905220032, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0488, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.24458816647529602, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0473, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.24380528926849365, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0411, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.1836114227771759, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0433, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.21051307022571564, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0462, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.18093563616275787, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.045, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2444952130317688, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0485, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.19429947435855865, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0487, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.2294744998216629, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0479, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.2147439569234848, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0492, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.2777695059776306, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0419, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22944948077201843, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0458, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.2822294235229492, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0429, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.20201396942138672, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0435, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.1961182802915573, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0441, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.1653209924697876, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0473, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.2082156240940094, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0508, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.199684277176857, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0466, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.20871016383171082, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0436, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.22781118750572205, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0504, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2652275562286377, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0467, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.20812813937664032, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0456, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2533996105194092, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0458, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.18139654397964478, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0441, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.27243369817733765, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0472, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.326709121465683, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0487, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.26698657870292664, + "learning_rate": 2.999617967180403e-05, + "loss": 0.05, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.4119804799556732, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0455, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.21175070106983185, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0462, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.22503416240215302, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.0468, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.254734605550766, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0471, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26713427901268005, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0439, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.2739362120628357, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0455, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.26221534609794617, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0451, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26752719283103943, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.23052464425563812, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0481, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.162479966878891, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0435, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.24716691672801971, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0459, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.14981329441070557, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0407, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.23326952755451202, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0459, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.2200203239917755, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.046, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.21501530706882477, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0433, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.22950918972492218, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0454, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.2042679488658905, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0444, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.16468611359596252, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0434, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.2180141657590866, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0447, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.1658746749162674, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.24921023845672607, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0398, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0445, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.25189560651779175, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0429, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.2614912688732147, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0427, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.2890806794166565, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0458, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.24215207993984222, + "learning_rate": 2.999336925698801e-05, + "loss": 0.04, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.1806204915046692, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0435, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.18884161114692688, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0422, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.2760990262031555, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0427, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.22660709917545319, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0459, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.24862384796142578, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.045, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.1911306381225586, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0433, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.2612020969390869, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0414, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.20945195853710175, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.043, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.2436099350452423, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0472, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.23788926005363464, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0455, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.20128341019153595, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0456, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.24002164602279663, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0414, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.2653037905693054, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.042, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.15122903883457184, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0395, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.19299687445163727, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0452, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.2534617781639099, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0445, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.29057401418685913, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0398, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.19744732975959778, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.041, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.25267109274864197, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0439, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2406993955373764, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0445, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.2561663091182709, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0465, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.2137032002210617, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0407, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.3002893030643463, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0433, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.32190272212028503, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0444, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.2076684534549713, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0432, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.31036099791526794, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0431, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.21386218070983887, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0442, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.19012001156806946, + "learning_rate": 2.998948184763285e-05, + "loss": 0.048, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.2310018688440323, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0466, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.1984279453754425, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0443, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.24101053178310394, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0446, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.17791084945201874, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0403, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.2016495168209076, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0404, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.19435155391693115, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.21197110414505005, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0398, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.22241441905498505, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.20825718343257904, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0434, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.24533888697624207, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.046, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.2455352544784546, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0412, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.18937505781650543, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0425, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.22356615960597992, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0418, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.23797748982906342, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0457, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.23130296170711517, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0427, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.20548853278160095, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0439, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19323743879795074, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0432, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.24339382350444794, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0412, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.18780124187469482, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0416, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.19353246688842773, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0461, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.21522872149944305, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0412, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.26847484707832336, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.3781932294368744, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0401, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.3219687342643738, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0423, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.1699237823486328, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0421, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.2174658179283142, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0398, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.2433294802904129, + "learning_rate": 2.998488804893431e-05, + "loss": 0.042, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.2116415947675705, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0405, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.22813065350055695, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0432, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.24252736568450928, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.048, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.2009168565273285, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0431, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.19379755854606628, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.23313777148723602, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0384, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.22056345641613007, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0459, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.3403800129890442, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0439, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.17649835348129272, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0419, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.20594412088394165, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0418, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.2901265323162079, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0405, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.22897225618362427, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0424, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.20961393415927887, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0438, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.28585514426231384, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0477, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2239968478679657, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0398, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.1651894897222519, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0405, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.1731943041086197, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0407, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.1708967089653015, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0417, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.17396977543830872, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0359, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.21327704191207886, + "learning_rate": 2.998095036004704e-05, + "loss": 0.045, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.15585461258888245, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0403, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.21774765849113464, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0403, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.1889292299747467, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0408, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.1639491617679596, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0425, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.23745876550674438, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.043, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19481611251831055, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0453, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.21001650393009186, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0429, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.2114192098379135, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0419, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.2506379783153534, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0392, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.21439281105995178, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0422, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.24763159453868866, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0433, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.1719428300857544, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0422, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.25943970680236816, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0425, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22226977348327637, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0448, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.2018195390701294, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0408, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.20869500935077667, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0443, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.23311494290828705, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.042, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.2501373589038849, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0382, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.2650911509990692, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0377, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.21386376023292542, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0405, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.27908873558044434, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0431, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.18731699883937836, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0402, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.2566622197628021, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0419, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.2418505847454071, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0406, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.199744313955307, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0424, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.20134931802749634, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.042, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.15607263147830963, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0453, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.22614715993404388, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0413, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.22810052335262299, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0433, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.16557516157627106, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0424, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.20254060626029968, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0394, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2068638950586319, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0408, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.27261367440223694, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0433, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.23216570913791656, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0445, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.2546656131744385, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0381, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.21829853951931, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0472, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.17109332978725433, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0396, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.1746513843536377, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0413, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.22411789000034332, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0373, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.21182285249233246, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.038, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.19430142641067505, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0419, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.22196920216083527, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0395, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.20695635676383972, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0407, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.1721169799566269, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0396, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.23676829040050507, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0396, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.3019765317440033, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0443, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.23958736658096313, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0402, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.17959944903850555, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0434, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.1563226878643036, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0397, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.16746941208839417, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0402, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.26673364639282227, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0411, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.23634585738182068, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0427, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.19724313914775848, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0408, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.2072090059518814, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0389, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.18574072420597076, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0427, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.23390066623687744, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0411, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.2137259542942047, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0395, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.21458280086517334, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0422, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.1838904619216919, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.043, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.1834891140460968, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0389, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.2012249082326889, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.27410435676574707, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0473, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.21607685089111328, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0405, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.17525449395179749, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.043, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.23137639462947845, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0414, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.20958152413368225, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0404, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.21453802287578583, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.043, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.2522047758102417, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.04, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.23595386743545532, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0396, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.23375281691551208, + "learning_rate": 2.996358559911323e-05, + "loss": 0.042, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.21628527343273163, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0434, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.2463727742433548, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0391, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.19919127225875854, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.043, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.28369203209877014, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0463, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.30222833156585693, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0397, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.19382458925247192, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0402, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2075910121202469, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0392, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.2503066658973694, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0427, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.18662676215171814, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0397, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.2432677149772644, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0425, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.19900213181972504, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0422, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.12742096185684204, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.20504114031791687, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0404, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.19127227365970612, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0425, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.18699967861175537, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0419, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.20074906945228577, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0381, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.24741096794605255, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0389, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.22948721051216125, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0435, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.17772004008293152, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0412, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.228739395737648, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0406, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.18312685191631317, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0401, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.21340692043304443, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0403, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.17110474407672882, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0388, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.19640640914440155, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.2386539727449417, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.043, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.20800922811031342, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0412, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1889747679233551, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0377, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.1801699548959732, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0402, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.20975525677204132, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0426, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.22626695036888123, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0404, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.2192756086587906, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0438, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.21307197213172913, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0447, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.28529244661331177, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0412, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.26367515325546265, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.044, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.29029133915901184, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.14019586145877838, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0427, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19176369905471802, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0408, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.19753597676753998, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0395, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1900634765625, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0427, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.21698379516601562, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0452, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.21722441911697388, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0417, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.1760135143995285, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0405, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18030588328838348, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0425, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.20307210087776184, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0416, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.19713939726352692, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0401, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.18719582259655, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.19560874998569489, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.04, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1633717119693756, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0424, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.20183435082435608, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0397, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.23784901201725006, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0438, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.17358708381652832, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0393, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18991878628730774, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0387, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.20455898344516754, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.042, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.23264537751674652, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0447, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2586922347545624, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0402, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.20447152853012085, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0406, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.1814388930797577, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0412, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2679784595966339, + "learning_rate": 2.994497917870469e-05, + "loss": 0.039, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.1818714141845703, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0423, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.2153957635164261, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0391, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.1858970820903778, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0394, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.23248426616191864, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.17587599158287048, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0367, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.18720309436321259, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.039, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.17602671682834625, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0393, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.15901489555835724, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0418, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.18207870423793793, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0401, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.22800953686237335, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0416, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.20031091570854187, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0417, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.24054989218711853, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0438, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.201461061835289, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0369, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.15697605907917023, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0397, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.16894802451133728, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0403, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2567543089389801, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0421, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.20297718048095703, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0396, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.17331846058368683, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.038, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.2214743047952652, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0405, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.2107590287923813, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0388, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.1768934726715088, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0364, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.16018562018871307, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0405, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.18402984738349915, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0425, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.21482129395008087, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0399, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.20554907619953156, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0388, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.2401154637336731, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0369, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.15072989463806152, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0359, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.1649719625711441, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0399, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.2386736422777176, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0403, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.2389005720615387, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0416, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.14930105209350586, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0372, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.18874907493591309, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0356, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.21466751396656036, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0391, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1697709560394287, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.24686385691165924, + "learning_rate": 2.993190369330068e-05, + "loss": 0.04, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.19733406603336334, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0417, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.3981468677520752, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0375, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.20926636457443237, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0384, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.18989330530166626, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0365, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17171184718608856, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0386, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.20474061369895935, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0413, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.19314105808734894, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0358, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.1794378012418747, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0409, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.17263059318065643, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0385, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.17863675951957703, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0355, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.20071984827518463, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0386, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.2025180608034134, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0409, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.21840184926986694, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0397, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.197452113032341, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0379, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.15370187163352966, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0379, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.22221392393112183, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0396, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.18536874651908875, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.18744619190692902, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0378, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.21197383105754852, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0399, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.29221004247665405, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0355, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.2411249279975891, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.039, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.3187369108200073, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.036, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.18096214532852173, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0372, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.2743191719055176, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0401, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.1864418387413025, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0384, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.15518276393413544, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.039, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.18089881539344788, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0359, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.2218618243932724, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0427, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.17051172256469727, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0401, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.15884852409362793, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0378, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1561235636472702, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0383, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.19711007177829742, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0409, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.18595683574676514, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.043, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.18613699078559875, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0357, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.1390007734298706, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0356, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.25603482127189636, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0377, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.1902502328157425, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0381, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.18781402707099915, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0393, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.20504912734031677, + "learning_rate": 2.991569760600352e-05, + "loss": 0.042, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.17359918355941772, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0373, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.2029414027929306, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0412, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.23485742509365082, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0398, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1454097479581833, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0394, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.22343166172504425, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0386, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.17754614353179932, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0358, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.14276635646820068, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0405, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.2062550187110901, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0429, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2093406766653061, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.038, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.14510402083396912, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0374, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.22726939618587494, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0373, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.14905233681201935, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0392, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.15828309953212738, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0361, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.990945968675917e-05, + "loss": 0.037, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.1867590844631195, + "learning_rate": 2.990900562792633e-05, + "loss": 0.038, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.18278446793556213, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0358, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.24628379940986633, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0393, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.2195759266614914, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0372, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.2397889941930771, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0352, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.1563243269920349, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0361, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.16278347373008728, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0366, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1776551604270935, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0369, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.19893717765808105, + "learning_rate": 2.990533239951709e-05, + "loss": 0.039, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.21403104066848755, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0378, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.19885864853858948, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0393, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.18425336480140686, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0404, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.1930677592754364, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0369, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18271560966968536, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0351, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.16518554091453552, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0367, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.17990152537822723, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0349, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.19625109434127808, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0368, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.1891077160835266, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0407, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.20162540674209595, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.17586444318294525, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0346, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.13146457076072693, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0338, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.16059468686580658, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0366, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.1724517047405243, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0395, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.1847251057624817, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0365, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.20104378461837769, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.037, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.2712373733520508, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0347, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.18014875054359436, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0388, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.17479945719242096, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0388, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.24176514148712158, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0387, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.2453809529542923, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0381, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2105131894350052, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0369, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1763688176870346, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0384, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.16223029792308807, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0348, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.173785999417305, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.038, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.23269419372081757, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0393, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.19176329672336578, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0399, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.1559731364250183, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0375, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.15701691806316376, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.22872254252433777, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0371, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15535064041614532, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0378, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.21613062918186188, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0382, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.1504523903131485, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0386, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.177171990275383, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2307550013065338, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0354, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.2215310037136078, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0371, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.23132716119289398, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0389, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.1937142014503479, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0373, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.23848143219947815, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0355, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.23107734322547913, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0359, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1455046832561493, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0404, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.21890339255332947, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0421, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.1897345334291458, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0357, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.2016463577747345, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0363, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.15850050747394562, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0391, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.18169008195400238, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0377, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.26095741987228394, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0391, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.20983633399009705, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0355, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.21026623249053955, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.2104070633649826, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0343, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1801508367061615, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0369, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.22245049476623535, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.0358, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.19220919907093048, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0354, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.2285826951265335, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0391, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.15382888913154602, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0374, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.13551367819309235, + "learning_rate": 2.987759237035982e-05, + "loss": 0.039, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.23362696170806885, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0367, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.15972478687763214, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0319, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.23858226835727692, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0354, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.17988596856594086, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0391, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.21262454986572266, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0392, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.1443355828523636, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0384, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.24133095145225525, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0382, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.175334632396698, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.039, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.20340824127197266, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0376, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.16549751162528992, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0371, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.16527467966079712, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0373, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1418737769126892, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0378, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18754976987838745, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0355, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.13649867475032806, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0329, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.21408453583717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0348, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.20929333567619324, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.24083013832569122, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0378, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17924576997756958, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0369, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.22360439598560333, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0382, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.15963956713676453, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0361, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.19003787636756897, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0365, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2458433210849762, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0375, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.22048939764499664, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0382, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.19965334236621857, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0371, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.228081613779068, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0375, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.24077889323234558, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0347, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.17692111432552338, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0365, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.1996857225894928, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0358, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.24799935519695282, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0377, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.24300269782543182, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0362, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.2230309098958969, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0364, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.16640354692935944, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0346, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.14670157432556152, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.039, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18531887233257294, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0353, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.19566190242767334, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0391, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.1967019885778427, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0355, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.2161232978105545, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0375, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.17630350589752197, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0397, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18905608355998993, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0347, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.1890578716993332, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.036, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.15397298336029053, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0401, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0362, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.20983757078647614, + "learning_rate": 2.985388676461079e-05, + "loss": 0.036, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.1554982215166092, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0378, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.1961488276720047, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0356, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.17265456914901733, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0357, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.1275556981563568, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0353, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.1667783409357071, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0354, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.27920690178871155, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0385, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.12451875954866409, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0352, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.15794622898101807, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0339, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.16285695135593414, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0369, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.18990163505077362, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0376, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.1867877095937729, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0363, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.15270273387432098, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0394, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.1934751272201538, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0382, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16943520307540894, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0369, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.20546840131282806, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0336, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.18490619957447052, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0415, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.16439616680145264, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0384, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.13623777031898499, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0379, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.18209043145179749, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0357, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.19922122359275818, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0365, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.15637168288230896, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0323, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1781233251094818, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0329, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.17547845840454102, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0344, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.1770639270544052, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0358, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.2677673101425171, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0356, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.16740617156028748, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0344, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.13076575100421906, + "learning_rate": 2.983793562696423e-05, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.17586535215377808, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0381, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.1623464673757553, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.037, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.16550037264823914, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0373, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.20314498245716095, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0366, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.19387134909629822, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0377, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.17531387507915497, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0381, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.21466659009456635, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.037, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1693795919418335, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0339, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.21929262578487396, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0332, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.2253660261631012, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0339, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.2084304839372635, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0363, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.2513981759548187, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0357, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.19585898518562317, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0353, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.22262930870056152, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0357, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.19813011586666107, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0347, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.16626323759555817, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0346, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.271424800157547, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.036, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2283792346715927, + "learning_rate": 2.982684510938631e-05, + "loss": 0.04, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.15996390581130981, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0394, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.20269536972045898, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0387, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.20575514435768127, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0386, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.15178649127483368, + "learning_rate": 2.982433099619666e-05, + "loss": 0.037, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.23187878727912903, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0355, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.12988153100013733, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0369, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19367633759975433, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0367, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.15319304168224335, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0337, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.19311629235744476, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0383, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.14771778881549835, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0393, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.20381954312324524, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0414, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22292324900627136, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0358, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.13758333027362823, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0359, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.18837112188339233, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0339, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.1757460981607437, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0352, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.24256370961666107, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0346, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.2417164295911789, + "learning_rate": 2.981603574229741e-05, + "loss": 0.037, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.21129392087459564, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0371, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16878797113895416, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0367, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.17526096105575562, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0353, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.1812993884086609, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.037, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.16877815127372742, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0353, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.20823006331920624, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0384, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1536571979522705, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0356, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.1477498710155487, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0346, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.12916848063468933, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0318, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.15418170392513275, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0369, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.1883322149515152, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0369, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.1569332480430603, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0374, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.21140694618225098, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0365, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.26859667897224426, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0395, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.2182101309299469, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0382, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.18896763026714325, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0372, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.17398814857006073, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0338, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.14043423533439636, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0361, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.14995983242988586, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0363, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.15542328357696533, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0361, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.16333894431591034, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0337, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.18489807844161987, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0352, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.1794673502445221, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.037, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.1882123351097107, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0414, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.16483181715011597, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.04, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.1727719008922577, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0361, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.1968923956155777, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0389, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1838955134153366, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.034, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.23556186258792877, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0369, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.2518751323223114, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0331, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18682284653186798, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0371, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.2049458622932434, + "learning_rate": 2.979480675356209e-05, + "loss": 0.04, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.11846950650215149, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0347, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.21428292989730835, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0354, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.154558002948761, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0359, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1178995817899704, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0364, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.16926616430282593, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0373, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.1721247136592865, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0346, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.1975203901529312, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0351, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.17340779304504395, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0347, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.22424420714378357, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0384, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.1688208281993866, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.16542011499404907, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0343, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.15097306668758392, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0372, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.15530438721179962, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0345, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.16840992867946625, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.034, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.19959625601768494, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0375, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.15717248618602753, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0344, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.21473488211631775, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.04, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.18634898960590363, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0346, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.16032545268535614, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0377, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.19153323769569397, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0408, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.16018801927566528, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0334, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.1498652696609497, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0349, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.1652575582265854, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0344, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.15636402368545532, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0331, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.22963644564151764, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0359, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.16931135952472687, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0389, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.17646029591560364, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0322, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.19198739528656006, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0318, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.16826699674129486, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0389, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.1693607121706009, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.037, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.147409588098526, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0343, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.1984914094209671, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0329, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.18371893465518951, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0357, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.16279958188533783, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0377, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2122160941362381, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0367, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.20099543035030365, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0343, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.17249372601509094, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0325, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1106863021850586, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0336, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.2338913530111313, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0331, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.1545400768518448, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0321, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.17114387452602386, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0323, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.16218030452728271, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.037, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.14618389308452606, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0306, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.17804935574531555, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0348, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1512002944946289, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0318, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.14877542853355408, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.037, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.1693819761276245, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0325, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.17713536322116852, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0356, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.1719064563512802, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0358, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.19448526203632355, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0363, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.18047156929969788, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0326, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.20896470546722412, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0362, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.15999653935432434, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0318, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.17288383841514587, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0336, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.21483922004699707, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0367, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.19133612513542175, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0336, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.16010481119155884, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0377, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.1778886914253235, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0354, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.14997199177742004, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0326, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.16331784427165985, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0326, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.18247577548027039, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0367, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1838751882314682, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0338, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.17194764316082, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0365, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.1354181170463562, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0315, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.19473528861999512, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0345, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14786818623542786, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0354, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.1931883543729782, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0345, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.159915491938591, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0357, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14373569190502167, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0309, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.1468040943145752, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0339, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.1592605859041214, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0374, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.17073556780815125, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0365, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.13732466101646423, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0331, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.18718381226062775, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0337, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.179783895611763, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0383, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.15795794129371643, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0368, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.13824236392974854, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0343, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.2081076204776764, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0385, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.1407512128353119, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0338, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.1619846224784851, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0325, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.184702068567276, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0353, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.18360818922519684, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0349, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.17683662474155426, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0332, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.17690211534500122, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0364, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.18250274658203125, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0336, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.12902002036571503, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0344, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.1598006635904312, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0332, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.17691342532634735, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.037, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.2213180959224701, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0324, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.14762577414512634, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0321, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.14543993771076202, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.034, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.18970227241516113, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0377, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.15049827098846436, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0341, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.1485714316368103, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0361, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1893187314271927, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0354, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.2345002293586731, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0336, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.24119973182678223, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0325, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.18577414751052856, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0337, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.1941320151090622, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0355, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.17646151781082153, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0342, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.13801687955856323, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0351, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.15124225616455078, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0337, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19114945828914642, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.033, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.23505449295043945, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0357, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.15116260945796967, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.15656375885009766, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0322, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.18982963263988495, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0368, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.18300476670265198, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0379, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.16485247015953064, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0335, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.19399121403694153, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0353, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.22944431006908417, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0343, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.16755999624729156, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0351, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.15779320895671844, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0318, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.2113935351371765, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0343, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.14698673784732819, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0332, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19958600401878357, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0357, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20250289142131805, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0334, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.1955910474061966, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0381, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.13143742084503174, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0321, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.15840105712413788, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0333, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.20588421821594238, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0347, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.1881185919046402, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0345, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.16714425384998322, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.032, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.16526015102863312, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0316, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.1865321844816208, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0351, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.19202300906181335, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0325, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.15551044046878815, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0339, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.15684190392494202, + "learning_rate": 2.969840338753429e-05, + "loss": 0.034, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.16806578636169434, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0319, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.1554284542798996, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0348, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1836974322795868, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0375, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.20003411173820496, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0364, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.1747797578573227, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0328, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.2204974740743637, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0331, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.17183512449264526, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0381, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.14398148655891418, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.03, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.17428989708423615, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0329, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.13864420354366302, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0319, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.15052364766597748, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0357, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.21513856947422028, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0349, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19239531457424164, + "learning_rate": 2.968758727655112e-05, + "loss": 0.0377, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.22284342348575592, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0358, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.17206035554409027, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0349, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.1991468220949173, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0316, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.13542494177818298, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0329, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.14473386108875275, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0346, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.1972355991601944, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0348, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.17447993159294128, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0337, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.14419333636760712, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0355, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.19937965273857117, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0381, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.18988190591335297, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0331, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.1552465558052063, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0333, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.1769426017999649, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0344, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.17652767896652222, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0354, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1598699688911438, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0354, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.17556554079055786, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0372, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.18424467742443085, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0372, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.1591036170721054, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.038, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.1973133534193039, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0332, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.139867901802063, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0336, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.17536301910877228, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0339, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16841425001621246, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0319, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.24087557196617126, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.1419876366853714, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0341, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0315, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.14541317522525787, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0326, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.18411535024642944, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0382, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1769087016582489, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0355, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.16459472477436066, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0354, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.1770380288362503, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.034, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.16219645738601685, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0316, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.18393918871879578, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0366, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.1436525285243988, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0318, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.18282611668109894, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0376, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.16712592542171478, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0367, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.13408221304416656, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0305, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.17664852738380432, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0312, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.13730520009994507, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0323, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1278112679719925, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0322, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.1742715984582901, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0339, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.15022823214530945, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0348, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.17816156148910522, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0348, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.18016202747821808, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0357, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.21158602833747864, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0331, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1747511923313141, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0301, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.16840165853500366, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0363, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.1507093459367752, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0352, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.2081453651189804, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.035, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.1459163874387741, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0331, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.17468400299549103, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0339, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.15824441611766815, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0309, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.19502659142017365, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0305, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.1502636969089508, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.032, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.13870884478092194, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0356, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1596919149160385, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0347, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.1830880492925644, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0359, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.13567551970481873, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0319, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.12907959520816803, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0327, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.17357082664966583, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.0308, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.20002517104148865, + "learning_rate": 2.963613172841619e-05, + "loss": 0.035, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.13852699100971222, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0353, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.15445692837238312, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.11687827110290527, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0319, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.15454575419425964, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0366, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.1916663944721222, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0327, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.16530223190784454, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0309, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17331530153751373, + "learning_rate": 2.962976980219581e-05, + "loss": 0.032, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.17225238680839539, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0323, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.16390499472618103, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0324, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.16459178924560547, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0315, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.1416798084974289, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0348, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.1546057015657425, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0329, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15206973254680634, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0304, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.18791860342025757, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.03, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.17669418454170227, + "learning_rate": 2.962243236871436e-05, + "loss": 0.034, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.1335674524307251, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0323, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.17903365194797516, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0327, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.17771181464195251, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0341, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1532796174287796, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0328, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.18777944147586823, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0343, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.2553282082080841, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0342, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.24115385115146637, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0347, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.18773671984672546, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0324, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.18168678879737854, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0355, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.14390279352664948, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0323, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.13260799646377563, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0336, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.16469483077526093, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0341, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.1618199497461319, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0305, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.15524879097938538, + "learning_rate": 2.960942086834339e-05, + "loss": 0.034, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.14689622819423676, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0352, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.17949312925338745, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0329, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.2376648485660553, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0326, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.1302678883075714, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0356, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.16316285729408264, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0318, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.14314667880535126, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0347, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.13162434101104736, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0335, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.13411720097064972, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0334, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.20853494107723236, + "learning_rate": 2.960094148810064e-05, + "loss": 0.036, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.15634609758853912, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0314, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.16109681129455566, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0323, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.15969422459602356, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0292, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.13126519322395325, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0338, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.13577662408351898, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0323, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13226218521595, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0307, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.12283602356910706, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0305, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16819486021995544, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0316, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.1443309634923935, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.031, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.17339679598808289, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0335, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.12366446852684021, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0319, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.18073806166648865, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0355, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.16230227053165436, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.033, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.15816937386989594, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0331, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.15911494195461273, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0296, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.19500406086444855, + "learning_rate": 2.958564529914256e-05, + "loss": 0.034, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.16553588211536407, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0348, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.1780945360660553, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0297, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16839291155338287, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0317, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.26763999462127686, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0335, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.1292608082294464, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0319, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19598962366580963, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0346, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.16073159873485565, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0317, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.2037622481584549, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0324, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.11946277320384979, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0329, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.12799547612667084, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0329, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.23773977160453796, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0327, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.19241386651992798, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0371, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.15968935191631317, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0335, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.1712830811738968, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0302, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.16464142501354218, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0346, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.2122460901737213, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0319, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.16571341454982758, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.033, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1610543429851532, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0347, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14235250651836395, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0322, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.14999081194400787, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0299, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.033, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.1448807269334793, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0307, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.20123368501663208, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0315, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.16191548109054565, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.031, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.1475522518157959, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.17617955803871155, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0323, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.20633190870285034, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0287, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.13562145829200745, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0347, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.15800844132900238, + "learning_rate": 2.955719840324779e-05, + "loss": 0.034, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.1652759164571762, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0291, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.11819476634263992, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0309, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.17562547326087952, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0327, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.1934351772069931, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0342, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.1565367877483368, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.033, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.21048930287361145, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0357, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.16762706637382507, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0341, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.1267809122800827, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0322, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.18461033701896667, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0357, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.13451576232910156, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0327, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.15478897094726562, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0311, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1461619883775711, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0314, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.1432616114616394, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0339, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.14061163365840912, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0337, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.1950581669807434, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0315, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.13942912220954895, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.031, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.15901322662830353, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0324, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.19862298667430878, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0322, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.15155215561389923, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0302, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16394968330860138, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0299, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.13225089013576508, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0305, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.18203529715538025, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0319, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.15323573350906372, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0346, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.18822363018989563, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.038, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1510896533727646, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0336, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.12274178862571716, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.032, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.1287696361541748, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0341, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.16619601845741272, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0321, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.12426833063364029, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0298, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.1751766800880432, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0341, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.13578996062278748, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0306, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.19475996494293213, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0314, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.13291390240192413, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0341, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.16193236410617828, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0325, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.16108311712741852, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0356, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.16233862936496735, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0303, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.14042457938194275, + "learning_rate": 2.951955450025455e-05, + "loss": 0.031, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.14922280609607697, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0336, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.1389254927635193, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0324, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.21042059361934662, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0331, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.18529878556728363, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0311, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1345708668231964, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0316, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.14794935286045074, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0367, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.18609392642974854, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0326, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.1762595772743225, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0312, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.15549533069133759, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0342, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.1910371482372284, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0316, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.12789981067180634, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0286, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.11457879841327667, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0303, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.12534750998020172, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0313, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.17089471220970154, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0318, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.16774669289588928, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0326, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.18693485856056213, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0356, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.20419354736804962, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15851624310016632, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0328, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1284184604883194, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.1148369237780571, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0291, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.15302640199661255, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.037, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.16082480549812317, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.031, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.16006684303283691, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0346, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14866025745868683, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0323, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.15490789711475372, + "learning_rate": 2.949326462724779e-05, + "loss": 0.033, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.18960629403591156, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0302, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.16329076886177063, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0338, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1700190156698227, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0326, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.14541998505592346, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0332, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.13809162378311157, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.16794940829277039, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0345, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1897762268781662, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0304, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2064688503742218, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0332, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.14193041622638702, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0332, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.15571711957454681, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0339, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.14264710247516632, + "learning_rate": 2.948147907679037e-05, + "loss": 0.036, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.1411963850259781, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0344, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.18891720473766327, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0308, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.15993206202983856, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0339, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.15609434247016907, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0315, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.19272638857364655, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0325, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.2836095690727234, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0323, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16382543742656708, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0335, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.18899215757846832, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0301, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.17764721810817719, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0322, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.1808568239212036, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0304, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.166172593832016, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0298, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.18262767791748047, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0305, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.17854458093643188, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.033, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.16440987586975098, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.033, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.19478020071983337, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0288, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.2605016231536865, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0318, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1850014477968216, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0313, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.1339086890220642, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.17828164994716644, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0315, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.1235261932015419, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0328, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.13704872131347656, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0322, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1296597719192505, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0315, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.16906750202178955, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0347, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14643864333629608, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.031, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.16102853417396545, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0315, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.17575663328170776, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0312, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1488381028175354, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0318, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.14629733562469482, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0333, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.17051763832569122, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.034, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.14636048674583435, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0315, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.19957765936851501, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.032, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.15027561783790588, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0342, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1580146700143814, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.032, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1452188491821289, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0302, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.1924576759338379, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0323, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.199595108628273, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0326, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.1557701975107193, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0303, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.21232250332832336, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0324, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.13840343058109283, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0321, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.14219941198825836, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0314, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1747855842113495, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0342, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1730964183807373, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0334, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.12938249111175537, + "learning_rate": 2.94341324778745e-05, + "loss": 0.029, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1782408058643341, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0299, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.943188093053675e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.12131186574697495, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0277, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.19141671061515808, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0307, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.2431664764881134, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0326, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15145368874073029, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0329, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.19781085848808289, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0332, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.19477561116218567, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0317, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.1818031370639801, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0352, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.13725785911083221, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0272, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.1874770075082779, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0322, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.18701452016830444, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0304, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.15173715353012085, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.032, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.1315767467021942, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0331, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.16627946496009827, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0299, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.2122701108455658, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0308, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.12093837559223175, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0303, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.2072226107120514, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0331, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.14395366609096527, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0323, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.18454021215438843, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0349, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1647966057062149, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0342, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1907777190208435, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0347, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.14438189566135406, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0295, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.16802345216274261, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0283, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.1488419622182846, + "learning_rate": 2.940567320758981e-05, + "loss": 0.033, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14881044626235962, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0318, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.18762606382369995, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0322, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.15426765382289886, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0314, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.17882433533668518, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0361, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.16833186149597168, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0319, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.16899487376213074, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0283, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.15301142632961273, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0325, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.13270439207553864, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0305, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.16603541374206543, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0347, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.21959784626960754, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0299, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.24568438529968262, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0339, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.14889934659004211, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0284, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.20354624092578888, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0309, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.18124797940254211, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0352, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.13871392607688904, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0311, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.18997471034526825, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0299, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.09903991222381592, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0304, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.13336266577243805, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0306, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.1564595103263855, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0295, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.17495550215244293, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.17048844695091248, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0266, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.1257491409778595, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0299, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.15244479477405548, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0341, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.15737923979759216, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.03, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.13226579129695892, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0268, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.21549297869205475, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0323, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.2102872133255005, + "learning_rate": 2.937416924082395e-05, + "loss": 0.029, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.14369380474090576, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0306, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.1390914022922516, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0273, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.13536471128463745, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0299, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.13321688771247864, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0298, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13358351588249207, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0309, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.1732599437236786, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0319, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.17455634474754333, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0314, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.1714484691619873, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0345, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.18120428919792175, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0293, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.21681863069534302, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0335, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.12099924683570862, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0299, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.13233523070812225, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0372, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.16708864271640778, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0305, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.2122814804315567, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0319, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14159013330936432, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0343, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.13396364450454712, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0308, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.17630934715270996, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0349, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.16417403519153595, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0316, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.16593796014785767, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0291, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.13613900542259216, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0299, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.15980054438114166, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0319, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1578991562128067, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0288, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.1628173291683197, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0317, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.15937024354934692, + "learning_rate": 2.934549758921277e-05, + "loss": 0.031, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.15258124470710754, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0323, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.16142596304416656, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0289, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.15643659234046936, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0295, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.17128989100456238, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0313, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.1871197372674942, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0314, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.1448175013065338, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0318, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.21338209509849548, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0334, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15262271463871002, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0293, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.14302000403404236, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0307, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.13756762444972992, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0288, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.1390128880739212, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0286, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.16088609397411346, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0299, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.11852050572633743, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0284, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.17093200981616974, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0331, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.2325763702392578, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0269, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.19948336482048035, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0314, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.16918544471263885, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0311, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.14563684165477753, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0301, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.19645430147647858, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0312, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.18950732052326202, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0292, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.11659497767686844, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0307, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1314595341682434, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0326, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1457567662000656, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.16525784134864807, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0322, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.15977199375629425, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0334, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.17081087827682495, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0308, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.20844970643520355, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.029, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.22166086733341217, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1766587346792221, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0298, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1307239532470703, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0298, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0289, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.17264464497566223, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0317, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.15076857805252075, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0308, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14443998038768768, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0291, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.13115015625953674, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0329, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1444655954837799, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0294, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.16740760207176208, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0318, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.14201150834560394, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0314, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.14225733280181885, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0297, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.16758595407009125, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0307, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.16751281917095184, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.03, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.1855877935886383, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0333, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.17420977354049683, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0277, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.1568649560213089, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0312, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.15874658524990082, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0297, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.13727585971355438, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0334, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.14999760687351227, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0328, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.16789689660072327, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.14890174567699432, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0312, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.18001733720302582, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0322, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.1718987226486206, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.03, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.18727312982082367, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0297, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.1528918594121933, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0296, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.13372628390789032, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0288, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.14980314671993256, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0319, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.11368083208799362, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0312, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.14968374371528625, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0311, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.15593667328357697, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0287, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.1762111783027649, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0296, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.15493011474609375, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0272, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.14142747223377228, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.03, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.1307222545146942, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0294, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.1598598062992096, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.03, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.17966923117637634, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0337, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15533174574375153, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0276, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.1223452091217041, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0294, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.12733329832553864, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0289, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.15903618931770325, + "learning_rate": 2.926085774098786e-05, + "loss": 0.029, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1539735347032547, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0272, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.20487286150455475, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0312, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.19920656085014343, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0315, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.15179695188999176, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0305, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.16460831463336945, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0315, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.21031928062438965, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0283, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.13835182785987854, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0287, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.1597253829240799, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0336, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.11488606780767441, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0328, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.18707934021949768, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0328, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.18874959647655487, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0317, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.1182878389954567, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0316, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.11695742607116699, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0308, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.1597018986940384, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0324, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.17729975283145905, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0316, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.18250048160552979, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0309, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.20568659901618958, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0293, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.12907297909259796, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0281, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.1809365600347519, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0333, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.17379151284694672, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0327, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.16082972288131714, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0308, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.17415443062782288, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0299, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.1659587323665619, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0272, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.1538572907447815, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0309, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.16303570568561554, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0322, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1444631963968277, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0306, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.24458004534244537, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0285, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.15384158492088318, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0277, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1425933837890625, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0303, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.12485234439373016, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0279, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.14287105202674866, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0297, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.1621903032064438, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0268, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.145540252327919, + "learning_rate": 2.921797581830774e-05, + "loss": 0.03, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.1410767287015915, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0309, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.10953503102064133, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0299, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.15029482543468475, + "learning_rate": 2.92140191009304e-05, + "loss": 0.031, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.13262124359607697, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0299, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1503380835056305, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0329, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.18932904303073883, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.028, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.1473148614168167, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0315, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.18515919148921967, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0309, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.16122235357761383, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0279, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.1684419959783554, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0309, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.21400189399719238, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0304, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.12552157044410706, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0286, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.15383073687553406, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.029, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.1269458532333374, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0289, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.1839684545993805, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0316, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.13441112637519836, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0281, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.13045959174633026, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0296, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.13933542370796204, + "learning_rate": 2.919408984114327e-05, + "loss": 0.029, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.15143601596355438, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0296, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.16731251776218414, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.032, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.13245129585266113, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0268, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.17244738340377808, + "learning_rate": 2.918873439576597e-05, + "loss": 0.028, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.1709858626127243, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0293, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18727441132068634, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0302, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1482084095478058, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0297, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.1343366652727127, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0303, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.1456393450498581, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0299, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1569104641675949, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0281, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.19711613655090332, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0298, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.14027468860149384, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.031, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.16852174699306488, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0289, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.18446287512779236, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0308, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.16207733750343323, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0283, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.19747264683246613, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0301, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.15206509828567505, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0295, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.18873688578605652, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0286, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.144051656126976, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0335, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.14620175957679749, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0284, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.19520114362239838, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0304, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.16127106547355652, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0333, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.11467311531305313, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0303, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11777719110250473, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0311, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.12151405215263367, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0278, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1852089911699295, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0297, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.19555963575839996, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0321, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.1584533303976059, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0292, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.1764054298400879, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0306, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.16257385909557343, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0298, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.13896049559116364, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0296, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.16509538888931274, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0321, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.14912301301956177, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0285, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.18058481812477112, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0302, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.18734610080718994, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0289, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.19707795977592468, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0327, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.16215947270393372, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0302, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1710345447063446, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0332, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.19638675451278687, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0316, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.17807847261428833, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0285, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.13119569420814514, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0298, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.13414712250232697, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0301, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.1941734254360199, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0327, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.19624412059783936, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0291, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.17615875601768494, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0293, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.19011878967285156, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0316, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.15080299973487854, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0263, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.1676572859287262, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0318, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.14975300431251526, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0263, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17691564559936523, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.17775340378284454, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0298, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.14438258111476898, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.14729741215705872, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0289, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.20924119651317596, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0297, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.15471352636814117, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0289, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.1571032702922821, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0307, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15048687160015106, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0301, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.15063059329986572, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0279, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15528838336467743, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0299, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.1756942719221115, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0291, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.1491880565881729, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0296, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.1562451720237732, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0302, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.12366753816604614, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0257, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.15308785438537598, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.16614912450313568, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0301, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.12428635358810425, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0303, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.135500967502594, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0283, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.16944879293441772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0279, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.174363911151886, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0295, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.16406473517417908, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0284, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.2144644856452942, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0277, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.22006124258041382, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0264, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.15261462330818176, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0298, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.15205760300159454, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.14101643860340118, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0305, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16015839576721191, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.17772945761680603, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0312, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16818812489509583, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0283, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1448637843132019, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0325, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.14702047407627106, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0303, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.15547925233840942, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0299, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.18527188897132874, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.029, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.1533391922712326, + "learning_rate": 2.907944011202334e-05, + "loss": 0.029, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.15355326235294342, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0286, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.1503538340330124, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0312, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.12572433054447174, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0277, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.13614071905612946, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0303, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.13363786041736603, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0292, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.13318102061748505, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.0289, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.14687903225421906, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0301, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.13144652545452118, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0271, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.16351011395454407, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0292, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.19256997108459473, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0268, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13006120920181274, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0285, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.12932874262332916, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.11981740593910217, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.14302268624305725, + "learning_rate": 2.905937392537971e-05, + "loss": 0.029, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.1446913331747055, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0289, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.1500432789325714, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.15604284405708313, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0283, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.1795264482498169, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0289, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.14356553554534912, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0304, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.1323058009147644, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0302, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.2098175585269928, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0297, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.1424631029367447, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0295, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.1956447958946228, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0294, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.11993541568517685, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0273, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.18414588272571564, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0316, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.17320625483989716, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0319, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.17650462687015533, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0304, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.18740026652812958, + "learning_rate": 2.903909846652783e-05, + "loss": 0.034, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.12898528575897217, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0309, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.11721221357584, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0297, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.1659565567970276, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0294, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.1469241976737976, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0298, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14588896930217743, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0332, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.16243143379688263, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0305, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.15857404470443726, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0318, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.14992372691631317, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0305, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.13760966062545776, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0299, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.14302772283554077, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0313, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.18560320138931274, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0262, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.13117150962352753, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0286, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.16370882093906403, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0298, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.18689852952957153, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.031, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.16412676870822906, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0289, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.20153313875198364, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0297, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.15255704522132874, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0316, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.2036312222480774, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0306, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.16087041795253754, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0315, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.1514595001935959, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0295, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.13961558043956757, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0259, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1996241807937622, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0336, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.15389376878738403, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0278, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.12345903366804123, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.032, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.16360828280448914, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0292, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.19773486256599426, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.032, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.11931311339139938, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0268, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.1378420740365982, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0296, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.09076640754938126, + "learning_rate": 2.899643488777395e-05, + "loss": 0.026, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.1116260215640068, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0266, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.18670536577701569, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0314, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.15351636707782745, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0279, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.15310993790626526, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0298, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.15458840131759644, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0329, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.12034284323453903, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0266, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.14272604882717133, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0296, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.16335424780845642, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0266, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.16105882823467255, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0355, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.1455371230840683, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0261, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.21555180847644806, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0301, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.15278705954551697, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0292, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.14099596440792084, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0287, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.10848120599985123, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.028, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.1380046308040619, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0273, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.14006251096725464, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0328, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.16518674790859222, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0286, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.13295377790927887, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.031, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.1623251736164093, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0281, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.16029980778694153, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.03, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.15575948357582092, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0294, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.157670795917511, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0294, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1470523178577423, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0281, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.12192463129758835, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0248, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.1240374892950058, + "learning_rate": 2.895893835994484e-05, + "loss": 0.028, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1385754942893982, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0287, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.140727236866951, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0281, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.15758906304836273, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0294, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.19664499163627625, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0277, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.12567110359668732, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0279, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.13595536351203918, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0313, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.14262230694293976, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0317, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1474338322877884, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0282, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.18601401150226593, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0274, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.1392786204814911, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0269, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.12022347748279572, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0262, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.17340584099292755, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0277, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.1438349187374115, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0273, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.1408676654100418, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0277, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.14853249490261078, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0291, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.20401965081691742, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.031, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.18385861814022064, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.12695786356925964, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0267, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.14115597307682037, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0291, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.15790492296218872, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0294, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.15104074776172638, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0278, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13686735928058624, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0276, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.12029985338449478, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0305, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1351509988307953, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0293, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.10609307885169983, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0285, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15056312084197998, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0292, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.17062142491340637, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0281, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.15565407276153564, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0298, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.136429563164711, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0282, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.1649797409772873, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0267, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.19815988838672638, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.032, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15660062432289124, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.03, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.1840171217918396, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0342, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.14063017070293427, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0282, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.14935986697673798, + "learning_rate": 2.890533053632388e-05, + "loss": 0.029, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.14691074192523956, + "learning_rate": 2.89037798520224e-05, + "loss": 0.033, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.14992092549800873, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0262, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.16885656118392944, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0301, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1223178282380104, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0275, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.1200852021574974, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0277, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.1251501888036728, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0252, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.16905754804611206, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0277, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.13312722742557526, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0296, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.13198228180408478, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0269, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.14326612651348114, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0253, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.14861920475959778, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0262, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.19368910789489746, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0295, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.143910750746727, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0282, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.12314137071371078, + "learning_rate": 2.888352491214935e-05, + "loss": 0.027, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.16095006465911865, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0272, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.1414606273174286, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.029, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.15631091594696045, + "learning_rate": 2.887882538107974e-05, + "loss": 0.0291, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16476406157016754, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0283, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.2236901968717575, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0286, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.1318763792514801, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0274, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.15661245584487915, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.029, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.13406990468502045, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0287, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16861270368099213, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0272, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.18082283437252045, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0285, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.17517460882663727, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0277, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.17011325061321259, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0325, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14378181099891663, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0306, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1766553819179535, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0304, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.1666528880596161, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0294, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.15178363025188446, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0273, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.14275510609149933, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0291, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.1568918377161026, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0267, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.1370164304971695, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0268, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.10388749092817307, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0289, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11448585987091064, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0302, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.15148931741714478, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0292, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.14388935267925262, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0257, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.14465516805648804, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0272, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.13414210081100464, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0273, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12774589657783508, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.028, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.24874050915241241, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0282, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.1624344289302826, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0294, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.12817884981632233, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0324, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1509743630886078, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0269, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.31345340609550476, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0287, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.1523321270942688, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0259, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.1633133739233017, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0269, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.14278745651245117, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.0253, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.18361203372478485, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0282, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.10244551301002502, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0268, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.125647634267807, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0264, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.12238023430109024, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0264, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.16535502672195435, + "learning_rate": 2.882169202177678e-05, + "loss": 0.0329, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.19624854624271393, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0303, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.18285983800888062, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0266, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.13460052013397217, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.0293, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.1427820920944214, + "learning_rate": 2.881525979398615e-05, + "loss": 0.0244, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.14855900406837463, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0276, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.1433439403772354, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0299, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.1452850103378296, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0271, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.14403793215751648, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0264, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.13841332495212555, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0277, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.1426801085472107, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0296, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.14492706954479218, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0272, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.1315297782421112, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.0289, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.1513436734676361, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0284, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.15784604847431183, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.0273, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.12729156017303467, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0297, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.20208381116390228, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0284, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.1486770659685135, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0271, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.1463368982076645, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.0292, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.12362723797559738, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0252, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.18984590470790863, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0292, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.12962038815021515, + "learning_rate": 2.878773564250086e-05, + "loss": 0.0301, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.132072314620018, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0298, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.1482372134923935, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0267, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.15687556564807892, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0308, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.11656040698289871, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.028, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.16482318937778473, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0256, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.13332504034042358, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.027, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.16453887522220612, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0253, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.16902299225330353, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0283, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.12483953684568405, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0276, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.14033488929271698, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0277, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.18922871351242065, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0258, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.15026360750198364, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0298, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17492659389972687, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0275, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.14793051779270172, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0301, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.17575892806053162, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.027, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.1661873608827591, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0275, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.14529843628406525, + "learning_rate": 2.875990888327069e-05, + "loss": 0.0277, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.11084084957838058, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0278, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1433780938386917, + "learning_rate": 2.875661528053842e-05, + "loss": 0.026, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.13195206224918365, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0298, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.14103750884532928, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0309, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.17067542672157288, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0296, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.14078323543071747, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0288, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.17906904220581055, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0268, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.163010373711586, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0285, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.17636623978614807, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0276, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.18726587295532227, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0297, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.18583180010318756, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.16773444414138794, + "learning_rate": 2.874008460348417e-05, + "loss": 0.0289, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.14579996466636658, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.12655268609523773, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0246, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.15124735236167908, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0297, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.16273140907287598, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0295, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14014655351638794, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0262, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12657561898231506, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0292, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.13813915848731995, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.0287, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.12874144315719604, + "learning_rate": 2.872678492608539e-05, + "loss": 0.026, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.12086840718984604, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0245, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.15290603041648865, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0297, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.13261260092258453, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0263, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.17371073365211487, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0268, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.17300288379192352, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0264, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.16441872715950012, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0279, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.17883789539337158, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.03, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.14519335329532623, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0266, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.18063998222351074, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0275, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.19818824529647827, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0256, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1695982664823532, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0274, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.16905854642391205, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0277, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.1852719783782959, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.03, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.14601513743400574, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0261, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.11720956116914749, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0279, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.14177089929580688, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0268, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.14933669567108154, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0292, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15183576941490173, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0285, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.1255270391702652, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.025, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.1467735320329666, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0258, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.13636574149131775, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0278, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.12642863392829895, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0254, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.09835807979106903, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0284, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.11784205585718155, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0285, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.14453290402889252, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0301, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.19569514691829681, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0264, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.13860125839710236, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.026, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.16651827096939087, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0283, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.13682647049427032, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0309, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.14632774889469147, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0275, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.1401532143354416, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0261, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.12643243372440338, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0268, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1152961477637291, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0259, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.14027607440948486, + "learning_rate": 2.866951776928e-05, + "loss": 0.0306, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.12379784137010574, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0272, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.15595580637454987, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0288, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.12823393940925598, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0263, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1481468379497528, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0283, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.12768998742103577, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0262, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.15771906077861786, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0267, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.18978171050548553, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0295, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.1386374831199646, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0277, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.1583528220653534, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0273, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.09999477118253708, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0253, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.14432361721992493, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0289, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.19001086056232452, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.0295, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.13931159675121307, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0257, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.1621151715517044, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0269, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.15239235758781433, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0275, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12619581818580627, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0273, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.14698320627212524, + "learning_rate": 2.864043385528954e-05, + "loss": 0.0275, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.14121954143047333, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0271, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.16203583776950836, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0265, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.15562862157821655, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0267, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.16438758373260498, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0285, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.13769859075546265, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0274, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.14908984303474426, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0256, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.14057780802249908, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.0261, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.16056638956069946, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.0267, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.12299710512161255, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0257, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.12815117835998535, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0284, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.14666494727134705, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0303, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.15829558670520782, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0243, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1499258577823639, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.029, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.16584458947181702, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0272, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1765419989824295, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0278, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.17066679894924164, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0281, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13183729350566864, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0255, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.12179457396268845, + "learning_rate": 2.86093128283797e-05, + "loss": 0.028, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.10134928673505783, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0275, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16393089294433594, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0278, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1360260546207428, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0297, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.16158054769039154, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0249, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.12037541717290878, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.0286, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.16677190363407135, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0277, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1442420333623886, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0268, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.1756538599729538, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0282, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.11580783128738403, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0282, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.11671322584152222, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0279, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.15163642168045044, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0297, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.1390339583158493, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0282, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.11784064769744873, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0275, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.15270282328128815, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0282, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.09030280262231827, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0281, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.15108083188533783, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0274, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.1403200924396515, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0265, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.14890284836292267, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0303, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.11579708009958267, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0283, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.14590834081172943, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0268, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.1513289511203766, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0264, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.13494084775447845, + "learning_rate": 2.857082133984191e-05, + "loss": 0.026, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.11738834530115128, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.029, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.1497882753610611, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0271, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.17350555956363678, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0277, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.16718043386936188, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0269, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.12658995389938354, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0267, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15880689024925232, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0256, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.1480095237493515, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0263, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.13779640197753906, + "learning_rate": 2.85567006866004e-05, + "loss": 0.0282, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1088612824678421, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0245, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.17420272529125214, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0267, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.12365500628948212, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.025, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.13261641561985016, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0256, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.10485266894102097, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0253, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14070947468280792, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0284, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.1179763451218605, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0241, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13687312602996826, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0284, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.17160922288894653, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0263, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.17487400770187378, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0275, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.11352817714214325, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1871306300163269, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0277, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.13434743881225586, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0278, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.16361933946609497, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0257, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.19457586109638214, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0277, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.16905367374420166, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0323, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.14845560491085052, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0248, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.12934044003486633, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.026, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.11046823114156723, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0267, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.14174290001392365, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0255, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.15194296836853027, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.0257, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.1532754749059677, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0269, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19864243268966675, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0288, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.16765300929546356, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0271, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.15602844953536987, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0288, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.1571321040391922, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0264, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1625281423330307, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0268, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.12808986008167267, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0307, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.12471310794353485, + "learning_rate": 2.850496147627652e-05, + "loss": 0.027, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.1521507054567337, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0289, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1389119178056717, + "learning_rate": 2.850136142292186e-05, + "loss": 0.024, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1338133066892624, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13214652240276337, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.0274, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.17002162337303162, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0277, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.1433296948671341, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0267, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.19547788798809052, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0287, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.1916888952255249, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0296, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.17188303172588348, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0262, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.14335697889328003, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.0269, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14298376441001892, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0278, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.12892583012580872, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0278, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.16246432065963745, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0255, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.1237327978014946, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0252, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.12249717861413956, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0238, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.16828545928001404, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0262, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.12834671139717102, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0287, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.11583935469388962, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.029, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.10479022562503815, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0244, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.18407097458839417, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0276, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.12898766994476318, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0277, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.13034385442733765, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0241, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.13197654485702515, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0269, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.14327986538410187, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0281, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.15228267014026642, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0269, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.11858481913805008, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0261, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.16075590252876282, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0284, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.1527116447687149, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0283, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.1280883252620697, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0294, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1348433941602707, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0251, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.15039700269699097, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0272, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.16682399809360504, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.0277, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.11187390983104706, + "learning_rate": 2.844503681416359e-05, + "loss": 0.0298, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.1675758808851242, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0267, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.16022244095802307, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0301, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.16192372143268585, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0277, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.13092738389968872, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.12416383624076843, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.028, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.12514744699001312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0252, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.1705007702112198, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.0274, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.14251261949539185, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0264, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.147093266248703, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0279, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.12747326493263245, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0265, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.14222867786884308, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0276, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.12574070692062378, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0262, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.12812504172325134, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0225, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.13585571944713593, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0267, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.12168630957603455, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0236, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.12048719823360443, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0225, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.15165774524211884, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0267, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.15395723283290863, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.026, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.14715734124183655, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0259, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.11370360106229782, + "learning_rate": 2.840817735753433e-05, + "loss": 0.0284, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1264869123697281, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.028, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.13726982474327087, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0268, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1450515240430832, + "learning_rate": 2.840261328154438e-05, + "loss": 0.0251, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.1328781247138977, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.025, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.15474216639995575, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.0292, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.1338219791650772, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0279, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.16368664801120758, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.027, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.18125605583190918, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.0252, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.127846822142601, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0252, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.17743080854415894, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0275, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.16253551840782166, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0287, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.14212867617607117, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0274, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.1380675882101059, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0263, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.12773451209068298, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.0257, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.1556098759174347, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0264, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.16690294444561005, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0296, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.11710768938064575, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0252, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.17441801726818085, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0318, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.10790088772773743, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0251, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.16537238657474518, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0266, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.1414048671722412, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0276, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.1096765398979187, + "learning_rate": 2.836716156259365e-05, + "loss": 0.0258, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.14450953900814056, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0278, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15733887255191803, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0278, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.15253902971744537, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0275, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.1410975307226181, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0249, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1242721900343895, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0251, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.15358713269233704, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0287, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.15856830775737762, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.029, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.1777517944574356, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0254, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.10823221504688263, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0262, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.144780695438385, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0286, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12900282442569733, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0243, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.15370796620845795, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0238, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.15801627933979034, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0259, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.19058606028556824, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0257, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14663439989089966, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0281, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.10609344393014908, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.0261, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.11403398960828781, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0291, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.11631747335195541, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0243, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.1325336992740631, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0266, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.11245301365852356, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.10258981585502625, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0252, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.15847361087799072, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0258, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.15267148613929749, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0246, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.15297843515872955, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0248, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.1767774075269699, + "learning_rate": 2.831995639121568e-05, + "loss": 0.0292, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.12696890532970428, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.025, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.17774799466133118, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0267, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.15603452920913696, + "learning_rate": 2.83142492407012e-05, + "loss": 0.0285, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.12731599807739258, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0272, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.13816066086292267, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0265, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.14216956496238708, + "learning_rate": 2.830853299002291e-05, + "loss": 0.026, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1427476406097412, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0263, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.14667609333992004, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0282, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.11785806715488434, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0276, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.125101238489151, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0255, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.10735481232404709, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0248, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.14256443083286285, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0286, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12700040638446808, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0275, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.11069332808256149, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0283, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.1438131332397461, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.0269, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.1901932805776596, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0274, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.17823027074337006, + "learning_rate": 2.828557706390495e-05, + "loss": 0.029, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.13579529523849487, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0267, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.1431250274181366, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0279, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.11313948035240173, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0274, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1669740527868271, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0288, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.12642791867256165, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0267, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.15632158517837524, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0256, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.1404937356710434, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0263, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.13466882705688477, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0249, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.13014933466911316, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0255, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.18688003718852997, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0267, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.14594615995883942, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0253, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.1708635836839676, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0244, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.16193392872810364, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0265, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.13249778747558594, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0286, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.13875705003738403, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0262, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.19137892127037048, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0246, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.15596376359462738, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0248, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.1577422320842743, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0247, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.14322742819786072, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0263, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.13651713728904724, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0261, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11288725584745407, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0252, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.09636299312114716, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0267, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.11729153245687485, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0275, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.14827747642993927, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0265, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.1446153074502945, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0257, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.1329014152288437, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0247, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.12726181745529175, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0259, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.13508957624435425, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0258, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.14538758993148804, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0264, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.10585593432188034, + "learning_rate": 2.822755216255228e-05, + "loss": 0.0276, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.17435809969902039, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0281, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.1504661589860916, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0259, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.1678706258535385, + "learning_rate": 2.8221699881734e-05, + "loss": 0.029, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.13508610427379608, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0261, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12610375881195068, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0237, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.12272591143846512, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0231, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.1211056336760521, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0252, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.13204888999462128, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0276, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.11982017755508423, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0245, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.17140068113803864, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0236, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.12660211324691772, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0264, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1636001467704773, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0266, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.16939589381217957, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0276, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.15193505585193634, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0291, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16562964022159576, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0283, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.13603557646274567, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.024, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.14359496533870697, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0257, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.14077764749526978, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0247, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.14604462683200836, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0277, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.1240646094083786, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0262, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13144497573375702, + "learning_rate": 2.818639656220465e-05, + "loss": 0.0255, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.1384861171245575, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0227, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.16475403308868408, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0269, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.14310070872306824, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0258, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1498526632785797, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0263, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.13084106147289276, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0268, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.12004188448190689, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0252, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.14093606173992157, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0262, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.11016195267438889, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.0249, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.11101678013801575, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0246, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.1335095763206482, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0274, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1479913741350174, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0237, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.11868918687105179, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0231, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12799321115016937, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.027, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.17011503875255585, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0271, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.15915420651435852, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0274, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.11501206457614899, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0254, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.1457805633544922, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0274, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.1501278430223465, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0264, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.13190113008022308, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0272, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.14392758905887604, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0235, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.1375834047794342, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0241, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.14287824928760529, + "learning_rate": 2.814280752477756e-05, + "loss": 0.0269, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12234462797641754, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0259, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.1336476355791092, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0242, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.11315035820007324, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0245, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.14132073521614075, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0293, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.12053677439689636, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.0266, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.10896297544240952, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0268, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.11100424081087112, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.026, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.11256317049264908, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0237, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.14322060346603394, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0255, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.1655588448047638, + "learning_rate": 2.812283453196085e-05, + "loss": 0.026, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.14714404940605164, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0261, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.12317050993442535, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0247, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12872155010700226, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0258, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.1666097790002823, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0261, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.13425199687480927, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.16131597757339478, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0295, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.13762447237968445, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0242, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.12883082032203674, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0274, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.15029405057430267, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0269, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.13236507773399353, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0303, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.12054237723350525, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0218, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.1471109539270401, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0242, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.11926044523715973, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0275, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.13649292290210724, + "learning_rate": 2.809470494867301e-05, + "loss": 0.0231, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.15434928238391876, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0283, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.11536315083503723, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0271, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.12312797456979752, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0249, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.12750110030174255, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0235, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.1568569540977478, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0277, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.11667264997959137, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0258, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.14598897099494934, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0266, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.21723008155822754, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0258, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.14233075082302094, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0244, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.14836373925209045, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0265, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.146131694316864, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0258, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.16575932502746582, + "learning_rate": 2.807043872632495e-05, + "loss": 0.025, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.15142853558063507, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0245, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.11762700974941254, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0253, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.09714391082525253, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0233, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.1586424708366394, + "learning_rate": 2.806231820878714e-05, + "loss": 0.023, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.14317330718040466, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.0271, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.10633813589811325, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.0244, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.19832657277584076, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0249, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.14428019523620605, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0283, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.18295790255069733, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.0257, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.16204482316970825, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0285, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.14742101728916168, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0239, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1435142159461975, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0253, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.2000540792942047, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.16311310231685638, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0253, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.1238052025437355, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0249, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.16693037748336792, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0269, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.12382039427757263, + "learning_rate": 2.803581696238626e-05, + "loss": 0.025, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.13866499066352844, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0259, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.12111294269561768, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0258, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.13717366755008698, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0266, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.1119573786854744, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0264, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.1390918791294098, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.12881262600421906, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0251, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.1354016363620758, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0256, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.1285468488931656, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0315, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.17634150385856628, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0254, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.15077067911624908, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.0268, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16464342176914215, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0265, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.3138365149497986, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0255, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.15954458713531494, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0272, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15878041088581085, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0267, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.17743942141532898, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0273, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.16661205887794495, + "learning_rate": 2.800297040373295e-05, + "loss": 0.0268, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.13536596298217773, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.0241, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.13258172571659088, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0266, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.1524200588464737, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0276, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.12779775261878967, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0245, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1008654311299324, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0222, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.14413924515247345, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0297, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.14558148384094238, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0252, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.15591812133789062, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0248, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.12246204912662506, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0276, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.13106316328048706, + "learning_rate": 2.7982312890369e-05, + "loss": 0.0265, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.12663032114505768, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0234, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.1167084351181984, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15868185460567474, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0251, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13669976592063904, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0291, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.15100552141666412, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0237, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14283764362335205, + "learning_rate": 2.796987104810136e-05, + "loss": 0.024, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.13055892288684845, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0255, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.16733981668949127, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0244, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.1716628074645996, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0257, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.11838851124048233, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0288, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.14557431638240814, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0238, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.14981132745742798, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0254, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.11149631440639496, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0256, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.20500904321670532, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0255, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.12128293514251709, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0254, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.12934274971485138, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0254, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.1355874240398407, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0267, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13485173881053925, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0259, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.1481594294309616, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0251, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.15145331621170044, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0243, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.16746139526367188, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0284, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.10760390013456345, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0268, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.1460299789905548, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0264, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.15936219692230225, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0232, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.1373136043548584, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.0255, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.14508306980133057, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0243, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.153093621134758, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0262, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.15373852849006653, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0238, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.14033931493759155, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0267, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.1424216777086258, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0252, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.11820772290229797, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0253, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.13179944455623627, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0243, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.18073157966136932, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0256, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.12908786535263062, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0254, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.18455258011817932, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0263, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.14844198524951935, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0248, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.14291873574256897, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0268, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.105900838971138, + "learning_rate": 2.79029165248073e-05, + "loss": 0.027, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.14083242416381836, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0244, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.14579042792320251, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0239, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.13705924153327942, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0254, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.17981873452663422, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0225, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.13812783360481262, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0278, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12511365115642548, + "learning_rate": 2.789025068651819e-05, + "loss": 0.027, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.13009224832057953, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0245, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.145152747631073, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0281, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.1607820838689804, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0254, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.09400778263807297, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0258, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.1608203500509262, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0263, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.14968201518058777, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0247, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.1333402544260025, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0272, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.17871612310409546, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0256, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.10688412934541702, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0264, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.154439315199852, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0259, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.1270124316215515, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0279, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.15214042365550995, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0259, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.14094167947769165, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0261, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.11419129371643066, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.025, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.1447947919368744, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0258, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.12327348440885544, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0246, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.11033894866704941, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0259, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.128106951713562, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0247, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.13206088542938232, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0254, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.13664662837982178, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0274, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1349778026342392, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0262, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.10689893364906311, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0262, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.16241611540317535, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0269, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14955222606658936, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0251, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.13519948720932007, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0231, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.15076202154159546, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0272, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.15367662906646729, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0255, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.15819156169891357, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0242, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.12844862043857574, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0259, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.13392682373523712, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0245, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.15134917199611664, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.0287, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.14139734208583832, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0254, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.1562367081642151, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0246, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.1298898160457611, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0252, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13048380613327026, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0252, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.12535928189754486, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0271, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.19301816821098328, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0245, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.12262707948684692, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0262, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1394580751657486, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0272, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.21376213431358337, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0243, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.15442071855068207, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0285, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.13063938915729523, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0277, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.15023048222064972, + "learning_rate": 2.779844963973707e-05, + "loss": 0.025, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.17523249983787537, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0239, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.1312434822320938, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0234, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13352124392986298, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0249, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.12799884378910065, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0257, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.12704412639141083, + "learning_rate": 2.778765821684581e-05, + "loss": 0.025, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.11119242757558823, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0248, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.1564178764820099, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0282, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.13500694930553436, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0279, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.1542324423789978, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0258, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13888850808143616, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0281, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.14011257886886597, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.14887000620365143, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0254, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.15670792758464813, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0289, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.1288663148880005, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0242, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.13191169500350952, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.027, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.1320677250623703, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0239, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.13418279588222504, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0234, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.11019188910722733, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0232, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.13963419198989868, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0235, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.1163327619433403, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.028, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.13508521020412445, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0275, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.11098913848400116, + "learning_rate": 2.775078589916535e-05, + "loss": 0.025, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.16348809003829956, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0259, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.12986604869365692, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0246, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1302310824394226, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0259, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.11797303706407547, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.0252, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.1427793651819229, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0233, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.13411074876785278, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.0257, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.1528121531009674, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.11762527376413345, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0238, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.13639862835407257, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0257, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.1356336623430252, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0269, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.16879825294017792, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.13352762162685394, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0236, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.20271515846252441, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0296, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.11791368573904037, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0241, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14179861545562744, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.0256, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12654612958431244, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0243, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12589605152606964, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0222, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.12714968621730804, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0278, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.13410630822181702, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0248, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.12528881430625916, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0258, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.11806205660104752, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0308, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.14317870140075684, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0244, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14099767804145813, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0228, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.15186408162117004, + "learning_rate": 2.769825463131276e-05, + "loss": 0.0252, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.18532416224479675, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0262, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.1445159614086151, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0265, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14576950669288635, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0253, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1777651309967041, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0229, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.16440069675445557, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0234, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.12858232855796814, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0259, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.14329349994659424, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0254, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.1265982836484909, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0235, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13783425092697144, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0228, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.1327633559703827, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0248, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14867408573627472, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0257, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.1556752771139145, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0239, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.16407394409179688, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0261, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.14142248034477234, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0265, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.11142456531524658, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0221, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.127807155251503, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0265, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.14716863632202148, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0241, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.1455392986536026, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.024, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1665087193250656, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0253, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.16937066614627838, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0292, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.13188236951828003, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0254, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.1010027602314949, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.0236, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.15787169337272644, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.0279, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.169437438249588, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0246, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.22705122828483582, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0227, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.17403557896614075, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.027, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.1287415474653244, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0253, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.13813741505146027, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0238, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.11655540764331818, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.13267700374126434, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.025, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.10447681695222855, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0239, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.11865647882223129, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.025, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.13245069980621338, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0256, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.11377252638339996, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0266, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.11403302848339081, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0252, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.1518944650888443, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0286, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.13505733013153076, + "learning_rate": 2.761618148370765e-05, + "loss": 0.0253, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.12230793386697769, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0247, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.1600245237350464, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0267, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.13100600242614746, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.11071992665529251, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0254, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.16208025813102722, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.0287, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.1652526706457138, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.027, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14042513072490692, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.1674395203590393, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0257, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.138246089220047, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0272, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13758982717990875, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0227, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.15503156185150146, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0259, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.15050673484802246, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0279, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.13952329754829407, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.024, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.14903564751148224, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0242, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.09984514862298965, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.09952808916568756, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0233, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.14514309167861938, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0255, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.12824438512325287, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0232, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.15507018566131592, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0251, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.12279994785785675, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0225, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.13601353764533997, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0249, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.20022229850292206, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.0263, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.1407146006822586, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0226, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.1350381076335907, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0248, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.13220156729221344, + "learning_rate": 2.755998359318217e-05, + "loss": 0.0251, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.11746657639741898, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0256, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.10933379083871841, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0235, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.1263892501592636, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0244, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.14062181115150452, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.0255, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.1430615335702896, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0254, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.19069550931453705, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0254, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.16568611562252045, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0252, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.15911205112934113, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.0265, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.17054520547389984, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0265, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.15547682344913483, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0237, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.12778308987617493, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0242, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.1590787172317505, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0232, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13945519924163818, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.023, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.14507925510406494, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0247, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14541085064411163, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0245, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.17666012048721313, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0222, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.16692468523979187, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0218, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.1295796036720276, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0242, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.12527599930763245, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0273, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.13334719836711884, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0253, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.1175302267074585, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0251, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.14461147785186768, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0261, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.15658612549304962, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0236, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.15802229940891266, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0236, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.14738211035728455, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0252, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.1639637053012848, + "learning_rate": 2.750090542643057e-05, + "loss": 0.027, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1573086827993393, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0264, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.13114139437675476, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0238, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.1112227514386177, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.027, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.12209689617156982, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.13741743564605713, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0239, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14444704353809357, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0263, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.13822096586227417, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0246, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.14305464923381805, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0237, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13539855182170868, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0264, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.13791607320308685, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0234, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.205885648727417, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0234, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.1034194752573967, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0214, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.15530072152614594, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0258, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.1535874903202057, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0226, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.19778263568878174, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0238, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.175798699259758, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0255, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.1803887039422989, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0238, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.1906212419271469, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0253, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.13742655515670776, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.026, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.12772122025489807, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0261, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13805702328681946, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0259, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.14081542193889618, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0241, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.10023088753223419, + "learning_rate": 2.744810891043728e-05, + "loss": 0.0246, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.10111884027719498, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0215, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.11498598754405975, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0217, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10912947356700897, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.0257, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.13553178310394287, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0266, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.18210133910179138, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0234, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.11901862919330597, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0248, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.14622537791728973, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0234, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.1641179323196411, + "learning_rate": 2.742962757296782e-05, + "loss": 0.026, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.14369463920593262, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.0236, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.14231407642364502, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0213, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.10929716378450394, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0216, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.13811182975769043, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0241, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.13219693303108215, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0219, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.13322590291500092, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0259, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.14716853201389313, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0255, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.11750922352075577, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0272, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.1430707424879074, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1722290813922882, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.025, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.15502606332302094, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.0231, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.09124620258808136, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0246, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.16296057403087616, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0277, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.10881830751895905, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0222, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.11321897804737091, + "learning_rate": 2.739481230505444e-05, + "loss": 0.026, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.1362120658159256, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.03, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.14329490065574646, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0252, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.4512016475200653, + "learning_rate": 2.738782381719405e-05, + "loss": 0.0238, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12559202313423157, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0263, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.12616541981697083, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0248, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.1516106277704239, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0234, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.12785211205482483, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0264, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16447393596172333, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.026, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.1648014783859253, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0264, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.14414788782596588, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0249, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.10982059687376022, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0254, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.12677313387393951, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0263, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.12355965375900269, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0223, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.15190085768699646, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0246, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.13910633325576782, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0231, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.11962535977363586, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0259, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.1566285789012909, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0256, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.12282316386699677, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0253, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.16674686968326569, + "learning_rate": 2.735040900116229e-05, + "loss": 0.025, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.14292193949222565, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0277, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.1504344791173935, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0268, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.16317704319953918, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0229, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.20625433325767517, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0242, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.1895458698272705, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.024, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.1987384408712387, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.0258, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.18668580055236816, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.1807073950767517, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0246, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.11927719414234161, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.028, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.11955022811889648, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0259, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.15261328220367432, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0268, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.15620580315589905, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0254, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.12000948935747147, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0234, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.1188279539346695, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.0248, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.13279424607753754, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0241, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.13379216194152832, + "learning_rate": 2.731275407491117e-05, + "loss": 0.026, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.12054166197776794, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0215, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.15941590070724487, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0279, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.18678002059459686, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0263, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.21807490289211273, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.0239, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.12243734300136566, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.028, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.14628900587558746, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0248, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.17471152544021606, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0257, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.11189062893390656, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0242, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.16700562834739685, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0252, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.1213584691286087, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0251, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.14496786892414093, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0258, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.17899028956890106, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0248, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13594108819961548, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0246, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.1365566849708557, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.021, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.10815171152353287, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0227, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.11390877515077591, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0209, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.1541927456855774, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0213, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.1342182159423828, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0251, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.1300208568572998, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0249, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14917100965976715, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0274, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.14349427819252014, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.0269, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10514543205499649, + "learning_rate": 2.72605878418951e-05, + "loss": 0.024, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.12946246564388275, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0234, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.1120578870177269, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.0246, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.14252731204032898, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.0287, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.0984109416604042, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0234, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.13058412075042725, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0235, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.1468561738729477, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0263, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.1297929584980011, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0224, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.15071524679660797, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0254, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.15211531519889832, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0239, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.13758648931980133, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0254, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.13972750306129456, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0239, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1762881577014923, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0235, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.13277511298656464, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.025, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.11352386325597763, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0225, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.1340995877981186, + "learning_rate": 2.722476145603374e-05, + "loss": 0.0232, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.16372360289096832, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0224, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.130090594291687, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0231, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.10481512546539307, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0229, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.1453929841518402, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0255, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.133439838886261, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0248, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.1380738615989685, + "learning_rate": 2.721037237880308e-05, + "loss": 0.023, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.1386265754699707, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0264, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.14298051595687866, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.0247, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.11756420135498047, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0222, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1197751984000206, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0216, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.2194000482559204, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0243, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.14157775044441223, + "learning_rate": 2.719594991888059e-05, + "loss": 0.023, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.14389759302139282, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0228, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.12442908436059952, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0267, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.1188371554017067, + "learning_rate": 2.718872618272836e-05, + "loss": 0.027, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11137005686759949, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0237, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.10907725989818573, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0251, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.150213360786438, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0229, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.2098330855369568, + "learning_rate": 2.717908157623383e-05, + "loss": 0.0245, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.12558938562870026, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0257, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.13659170269966125, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0247, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.12037219852209091, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0228, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.13085514307022095, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0252, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.13916544616222382, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0251, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.13098886609077454, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.021, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.12777045369148254, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0258, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.10204357653856277, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0249, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.11189272999763489, + "learning_rate": 2.715732712163776e-05, + "loss": 0.0232, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.1412421613931656, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0255, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.09470678865909576, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0217, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.09736761450767517, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0233, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12988215684890747, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.1080017015337944, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0263, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.12020932137966156, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0238, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.1058862954378128, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0247, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11686211824417114, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.024, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.11078432947397232, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0226, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.13394472002983093, + "learning_rate": 2.713306780076029e-05, + "loss": 0.0237, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.11299733072519302, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.025, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.1196199283003807, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0251, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.16477037966251373, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0241, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.1175335943698883, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.11024156212806702, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0237, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.12922869622707367, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0278, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.1429518610239029, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0254, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.14210145175457, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0225, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.12270870059728622, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0257, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.12436466664075851, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0258, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.10704488307237625, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0224, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.15890415012836456, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0252, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.12372975796461105, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0225, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.11652565747499466, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.028, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.1483478546142578, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0255, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.13188686966896057, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0238, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.18725869059562683, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0244, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.15949182212352753, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0225, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.14752954244613647, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0229, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.14210951328277588, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0233, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.1388787031173706, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.0253, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.11010867357254028, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0227, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.13336122035980225, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0227, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.10118173807859421, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.0265, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.1434611827135086, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0263, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.14133326709270477, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.026, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.15359432995319366, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0236, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.12927602231502533, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.022, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.10330940783023834, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0236, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.1264553815126419, + "learning_rate": 2.705973772304842e-05, + "loss": 0.025, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.15007789433002472, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0267, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.11090201884508133, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0257, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.12769529223442078, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0252, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.0977141410112381, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0244, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.14598220586776733, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0254, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.11231301724910736, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0235, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.09610529243946075, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0262, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.1549275815486908, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0234, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.11284032464027405, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.0217, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.14465086162090302, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.0235, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.13792403042316437, + "learning_rate": 2.703264323632037e-05, + "loss": 0.024, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.1438492238521576, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0239, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.12613703310489655, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0244, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.14072397351264954, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0225, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.1420842856168747, + "learning_rate": 2.702276326859174e-05, + "loss": 0.026, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.17447218298912048, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0243, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.1708984673023224, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0259, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.1555069535970688, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.0252, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.12473946064710617, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0227, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.15750949084758759, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0248, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.14009647071361542, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0267, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.1416911780834198, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0229, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.11944608390331268, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0235, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.126204252243042, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0268, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.1186814159154892, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0242, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.12699876725673676, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0239, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.10452742129564285, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0221, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.1422647386789322, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0247, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.11601094901561737, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0232, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12339281290769577, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0223, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1883407086133957, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0243, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.1730756312608719, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0294, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.10429747402667999, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0194, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.1358909010887146, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0217, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.14417894184589386, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0265, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.12299799919128418, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0236, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.1361788511276245, + "learning_rate": 2.696816263672279e-05, + "loss": 0.024, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.14492252469062805, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0241, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.1209225058555603, + "learning_rate": 2.696317709902802e-05, + "loss": 0.0239, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.11048539727926254, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.12423776835203171, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0243, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.1254356950521469, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.12912537157535553, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0254, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.10864657163619995, + "learning_rate": 2.695069735893089e-05, + "loss": 0.022, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.15519478917121887, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0253, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.15276901423931122, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0231, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.1401076465845108, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0271, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.14237190783023834, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.024, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1186559721827507, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0238, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.1254940629005432, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0233, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.1337885558605194, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0255, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12958411872386932, + "learning_rate": 2.69306825906724e-05, + "loss": 0.025, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.11610005795955658, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0238, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.13515442609786987, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0239, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1492868959903717, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0224, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.1293300986289978, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.021, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12354710698127747, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.13226614892482758, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.0247, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.1368253529071808, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0255, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.13044875860214233, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0228, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14107753336429596, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0221, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.13521252572536469, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0236, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.13718804717063904, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0217, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.15192179381847382, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.0247, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.13718710839748383, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0244, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11781664192676544, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0239, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.14431379735469818, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0248, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.15059512853622437, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0231, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.12076479196548462, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0242, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.19006308913230896, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0238, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.140592560172081, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0233, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.1528412401676178, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0229, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.16418038308620453, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0242, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.12185576558113098, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0254, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11594245582818985, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0248, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1216709092259407, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0234, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.12024594098329544, + "learning_rate": 2.686776313743295e-05, + "loss": 0.024, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.1321716010570526, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.0255, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.13785192370414734, + "learning_rate": 2.686270520700719e-05, + "loss": 0.025, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.12321551889181137, + "learning_rate": 2.6860174890357e-05, + "loss": 0.0233, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14606279134750366, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0249, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.17512187361717224, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0272, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.14974310994148254, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.0243, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.11758870631456375, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0241, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.11157143115997314, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0232, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.10899091511964798, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0231, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.1136092096567154, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0224, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.10306857526302338, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0211, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13241226971149445, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0222, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.11886361986398697, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0244, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.11962216347455978, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0237, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.14516247808933258, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.14592237770557404, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0253, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1278424859046936, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0233, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.12451537698507309, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0218, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.11746378242969513, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0266, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.14618739485740662, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0226, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.1512313336133957, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0233, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.14035750925540924, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0264, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.13232699036598206, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.023, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.14616237580776215, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.0233, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.12882372736930847, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0265, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.13778620958328247, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.024, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.15017054975032806, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0239, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.14781621098518372, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0215, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1670294851064682, + "learning_rate": 2.679407107451536e-05, + "loss": 0.022, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.1112075075507164, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0232, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.12219138443470001, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0202, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12922540307044983, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.11967732757329941, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0226, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.1377563774585724, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0224, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.14771930873394012, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0234, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.14800170063972473, + "learning_rate": 2.677617030053916e-05, + "loss": 0.0251, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.1233784481883049, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0232, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.16905057430267334, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0214, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.13751909136772156, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0248, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.13289029896259308, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0216, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.10788674652576447, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0216, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.13367260992527008, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1270209699869156, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0239, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.140412375330925, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0218, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.17788565158843994, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0253, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.13695330917835236, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.023, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.1250872015953064, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0232, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.11347977817058563, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0243, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.15546919405460358, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0233, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.13752058148384094, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.10431966185569763, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0215, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.17878347635269165, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0262, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.13218404352664948, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0241, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.12244930118322372, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.0247, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.11582386493682861, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0219, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.1160157322883606, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0203, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.10731764137744904, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0234, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.11152671277523041, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0218, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.12760375440120697, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0218, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.13706916570663452, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0219, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.14939993619918823, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0208, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.1169784739613533, + "learning_rate": 2.67092985767871e-05, + "loss": 0.0217, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.1569128930568695, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12233320623636246, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0251, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13152818381786346, + "learning_rate": 2.670154386126402e-05, + "loss": 0.0231, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.1529632955789566, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0246, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.11001774668693542, + "learning_rate": 2.669636960731296e-05, + "loss": 0.0223, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.1473267376422882, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.025, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.12853027880191803, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0246, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.15698745846748352, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0209, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12175270169973373, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0229, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.15634675323963165, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.0224, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.16060872375965118, + "learning_rate": 2.668082553346281e-05, + "loss": 0.0254, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.13773131370544434, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0224, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.14258436858654022, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0216, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.13032476603984833, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0215, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.17300327122211456, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0221, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.13187018036842346, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0226, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.14065369963645935, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0232, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.11484711617231369, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0263, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1380661576986313, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0209, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.17800134420394897, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0244, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.14836716651916504, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0277, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14563506841659546, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.023, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.11463280767202377, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.026, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.13092686235904694, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0222, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.12265385687351227, + "learning_rate": 2.664443191002912e-05, + "loss": 0.025, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.14121955633163452, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0242, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.12938837707042694, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0222, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.1922816038131714, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.0228, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.12596914172172546, + "learning_rate": 2.663400187275054e-05, + "loss": 0.022, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.14095240831375122, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0244, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.1080891340970993, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0228, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.13916714489459991, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.0239, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.1489475965499878, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0251, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.11781706660985947, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0207, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.11544635146856308, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0232, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.14827974140644073, + "learning_rate": 2.661571529941175e-05, + "loss": 0.0236, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.12821446359157562, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.0186, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.1303764134645462, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0234, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.1049366369843483, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0232, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.15925322473049164, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0275, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.13151033222675323, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0213, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.09425143152475357, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0215, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.15108847618103027, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0228, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.1374080330133438, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0231, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.16061817109584808, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0253, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.1707434207201004, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0259, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.12051897495985031, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0217, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.09191743284463882, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0224, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.12547235190868378, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0245, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14491257071495056, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0228, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.11896242201328278, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0239, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.14253611862659454, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0233, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.12803201377391815, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.0226, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.12893305718898773, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0231, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.16127154231071472, + "learning_rate": 2.656586262343925e-05, + "loss": 0.0234, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.13307608664035797, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0258, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.11211924254894257, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.0219, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.1672789305448532, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0248, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.12878142297267914, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0231, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.11429315060377121, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0233, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.13469499349594116, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0254, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.12753085792064667, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0258, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.11875704675912857, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0234, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.13718342781066895, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0223, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.10762278735637665, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.021, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.11651597917079926, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0222, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.12953802943229675, + "learning_rate": 2.653421326308085e-05, + "loss": 0.0235, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.14820824563503265, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.023, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.16258545219898224, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0266, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.14097236096858978, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0224, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.1039188802242279, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0216, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15014079213142395, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0246, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12850438058376312, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.023, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.12245986610651016, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0209, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.15426407754421234, + "learning_rate": 2.651304359235998e-05, + "loss": 0.025, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.13411583006381989, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0229, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.13705122470855713, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0254, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.13120216131210327, + "learning_rate": 2.650509053370553e-05, + "loss": 0.022, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14599482715129852, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0249, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.11628340184688568, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0215, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.13643144071102142, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0228, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.12846967577934265, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0214, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.13181492686271667, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0245, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1277213841676712, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0244, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12288910150527954, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0226, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.14789454638957977, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0215, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.11457780748605728, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0248, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.12511487305164337, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0206, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.12255935370922089, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0245, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.13214053213596344, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0238, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.1601077914237976, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.027, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.1244364008307457, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0212, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.11869437992572784, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0239, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.14327670633792877, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0227, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.12729141116142273, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0248, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.10953658074140549, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0205, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.20326738059520721, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0225, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.15614672005176544, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0237, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.11863692849874496, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0208, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.10683806240558624, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0224, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.1529647260904312, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.024, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.18428578972816467, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0232, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.19935141503810883, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.0223, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.14675194025039673, + "learning_rate": 2.643583521046136e-05, + "loss": 0.0216, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.21206484735012054, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0243, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.14521193504333496, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0214, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.11565426737070084, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0236, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.1087750792503357, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0215, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.11942578107118607, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0229, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.138235405087471, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0219, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.1079302579164505, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0222, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.15928085148334503, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0225, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.10920153558254242, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.024, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.12247785925865173, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0205, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.10454121232032776, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0238, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.1726839691400528, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0211, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1617811918258667, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0225, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.15452679991722107, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0214, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1573309749364853, + "learning_rate": 2.639561293166059e-05, + "loss": 0.022, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.13596323132514954, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0234, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.09188699722290039, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0219, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.12158972024917603, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0202, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.16615170240402222, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.0221, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.1163380891084671, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0255, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.17704762518405914, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0243, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.1591942012310028, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.021, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.14697043597698212, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0241, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1234351247549057, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0261, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.1379440575838089, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0266, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1025586649775505, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.022, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.11894206702709198, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0226, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1431341916322708, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0241, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.10291697829961777, + "learning_rate": 2.635789644394928e-05, + "loss": 0.022, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.1622016727924347, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0229, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.11867184937000275, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0262, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.18177814781665802, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0237, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.15159954130649567, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0218, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.11481288820505142, + "learning_rate": 2.634438526911474e-05, + "loss": 0.024, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.10989677160978317, + "learning_rate": 2.634168044873305e-05, + "loss": 0.024, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.12370797991752625, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0235, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1390318125486374, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0231, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.14030151069164276, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.0207, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11659140884876251, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.021, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.16727864742279053, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0242, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.11524021625518799, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0265, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.10776443779468536, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0224, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.16366012394428253, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0264, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.1102370023727417, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0245, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11543556302785873, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0216, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.11101235449314117, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0208, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.1257167011499405, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.024, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.1063276082277298, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0241, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.14303457736968994, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0242, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.12862786650657654, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0244, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15487778186798096, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.14244592189788818, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0211, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1066693440079689, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0199, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11721482872962952, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0214, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.13803230226039886, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.0227, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.11225558817386627, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.028, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.11660541594028473, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0239, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12551093101501465, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0221, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.1305139809846878, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.1281510442495346, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0235, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.14735114574432373, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0213, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.10699455440044403, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0231, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.16579580307006836, + "learning_rate": 2.626559653214893e-05, + "loss": 0.0213, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.09389153122901917, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0218, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.17565062642097473, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0245, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.1368194818496704, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0214, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.10599412024021149, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0221, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.12645669281482697, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0233, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.12325828522443771, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.0228, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.11353788524866104, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.10889376699924469, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0227, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.12935040891170502, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0214, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.1422247439622879, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0219, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.16723784804344177, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0239, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.09775881469249725, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.0229, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.13455381989479065, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0253, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.13700921833515167, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0239, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12252738326787949, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0231, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.15983854234218597, + "learning_rate": 2.622181856488204e-05, + "loss": 0.0223, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.18716242909431458, + "learning_rate": 2.621907518918971e-05, + "loss": 0.025, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.16380536556243896, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0229, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.1524866819381714, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0221, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.12281060218811035, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0213, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.09793791174888611, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0249, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.12271476536989212, + "learning_rate": 2.620534553467634e-05, + "loss": 0.0221, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.12616175413131714, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0236, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.10274055600166321, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0205, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.16873067617416382, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.027, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.10157257318496704, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0236, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.11863094568252563, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0225, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.09981630742549896, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0233, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11045375466346741, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0226, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.13111108541488647, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0232, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.12417030334472656, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0226, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.12029916048049927, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0217, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.10766672343015671, + "learning_rate": 2.617506544598029e-05, + "loss": 0.025, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.12039721757173538, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0228, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.11044551432132721, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0223, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.12463348358869553, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.0232, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.14382848143577576, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0238, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.13677646219730377, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0239, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.11536712944507599, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0224, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.12461430579423904, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0253, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.16901540756225586, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0242, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.12255802005529404, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0212, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.14185196161270142, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.0218, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.1292722374200821, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0249, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12001507729291916, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0222, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.12286248803138733, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0211, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.12487491965293884, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0231, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13417202234268188, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0257, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.11690660566091537, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0219, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15474916994571686, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0223, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.1720103919506073, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0239, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.16786889731884003, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0216, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.2279045283794403, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.0254, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13872499763965607, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0238, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.1881643831729889, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0211, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.1372661143541336, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0237, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.10514581203460693, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0217, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.11103228479623795, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0248, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1349911093711853, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0237, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.12774604558944702, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0245, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.15125125646591187, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0233, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.1190832257270813, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0248, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.1088283434510231, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0187, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.12167634069919586, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.023, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.15135347843170166, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0226, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.12857669591903687, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.0195, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.13588359951972961, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0234, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.14902308583259583, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0235, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.15095068514347076, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0209, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.1116727739572525, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0255, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11652971059083939, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0211, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.11875079572200775, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0223, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.11100098490715027, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.022, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.13648055493831635, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0234, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.12870723009109497, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0205, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.10456091910600662, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0232, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.13770641386508942, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0212, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.1526155173778534, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0244, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12664297223091125, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0241, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.12000498175621033, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0239, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.11930232495069504, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.0207, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.11384909600019455, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0217, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.1746685951948166, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.023, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.15868088603019714, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0233, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.15131928026676178, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0194, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.10279352962970734, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0215, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.13810355961322784, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0202, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.13024795055389404, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0217, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1374605894088745, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0215, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.14356477558612823, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0206, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.11046172678470612, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0236, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.10146678239107132, + "learning_rate": 2.601090850395125e-05, + "loss": 0.021, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13247323036193848, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0217, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.13672757148742676, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0201, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.08651759475469589, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.0196, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.14821568131446838, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0223, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11985234171152115, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0208, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11231720447540283, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0219, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.12338588386774063, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0202, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.15079368650913239, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0229, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.1348753422498703, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0209, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1333470344543457, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0207, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.10075798630714417, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0232, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1217522993683815, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0247, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.12041056901216507, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.022, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.10997799783945084, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0226, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.119750015437603, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0239, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.15246307849884033, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0239, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.1431925892829895, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0227, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.1181643158197403, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0227, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.13645906746387482, + "learning_rate": 2.595742267991401e-05, + "loss": 0.0213, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.12661530077457428, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0233, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.1344967484474182, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.0221, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.13433857262134552, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0238, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.10517163574695587, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0237, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.10689525306224823, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.022, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.13307900726795197, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0209, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.14696283638477325, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0261, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.12473653256893158, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0213, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.106630340218544, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0221, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.12092334777116776, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0232, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.11033511906862259, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0193, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11916322261095047, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0212, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.1204444020986557, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0228, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.1373489499092102, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0225, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.13124211132526398, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0208, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.18858365714550018, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0258, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.1558879315853119, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0228, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.1481737494468689, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0242, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.10424906760454178, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.021, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.15494640171527863, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0238, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.1801268309354782, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0238, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.14176428318023682, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0228, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.15850430727005005, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0235, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.14727924764156342, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0232, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.14907293021678925, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0242, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.11011895537376404, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0231, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.13541805744171143, + "learning_rate": 2.588090010443142e-05, + "loss": 0.0224, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.13078846037387848, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0228, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.14508767426013947, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0207, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.12445177882909775, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0227, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.14860141277313232, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0244, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.13311992585659027, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0207, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.1157015860080719, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0209, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.13310302793979645, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.0234, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.14026974141597748, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0246, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.14788632094860077, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0213, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.1551811844110489, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0241, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.10706475377082825, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.021, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.12536053359508514, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0228, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.1260751187801361, + "learning_rate": 2.584384079351628e-05, + "loss": 0.022, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.1082935482263565, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0207, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.13922622799873352, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0251, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1240227147936821, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0223, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.14512665569782257, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0218, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12859801948070526, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.0208, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.13418391346931458, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.021, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.10546918958425522, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.022, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.10668392479419708, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10407276451587677, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.02, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.13980497419834137, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0235, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.16233955323696136, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0224, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.16743852198123932, + "learning_rate": 2.580950866038058e-05, + "loss": 0.022, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.09152300655841827, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0206, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.14319349825382233, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0205, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.1045396700501442, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0219, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.10865644365549088, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0201, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.12166012078523636, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0217, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.12505589425563812, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0224, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.1353408247232437, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0242, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.1379590630531311, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0221, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12337373197078705, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0235, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.12092723697423935, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0233, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.09944337606430054, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0215, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.15161453187465668, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0193, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.0963459238409996, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0225, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.15534570813179016, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0247, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.142191082239151, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0228, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.1454230695962906, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0226, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.12661893665790558, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0211, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.11373333632946014, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0213, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12031935155391693, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.0208, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.13401862978935242, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0224, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.11285468935966492, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0231, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.12453637272119522, + "learning_rate": 2.574625966768285e-05, + "loss": 0.0211, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.13041555881500244, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1391122043132782, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0224, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.1345762014389038, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0244, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.12814867496490479, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0219, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.11058632284402847, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0191, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.15668845176696777, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.021, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.1509203314781189, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.0254, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.13209059834480286, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0201, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.1357070952653885, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0226, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.0896834284067154, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0207, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.1570383608341217, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0234, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.16916486620903015, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0254, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.1385425180196762, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.0218, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.11236828565597534, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0214, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11224918812513351, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0233, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.1305687576532364, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0219, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.15051157772541046, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0226, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.13418416678905487, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0244, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.14324583113193512, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0235, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.10810746997594833, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0197, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.14513953030109406, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0215, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1187531128525734, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0193, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11865831911563873, + "learning_rate": 2.567971342427169e-05, + "loss": 0.021, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.12040061503648758, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0229, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.10597001016139984, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0209, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.1219123974442482, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0242, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.11732123047113419, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.0203, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.15642790496349335, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0226, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.1893257051706314, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0247, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11829031258821487, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0239, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.14971555769443512, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0253, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.12424097210168839, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0251, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1071329414844513, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0271, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.11146610230207443, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0214, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.10925749689340591, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.0212, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.14797845482826233, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0221, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.10348086059093475, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0218, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.14607182145118713, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0217, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.11097954958677292, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0259, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.11300431936979294, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0197, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.11342521011829376, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0202, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.12889035046100616, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0217, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.10449834913015366, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0216, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1293669193983078, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0227, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.10667148977518082, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.021, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.0839434340596199, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0236, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.12906265258789062, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0223, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.1100582480430603, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0243, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.10644520074129105, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.0215, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.11693669110536575, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.022, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.13665053248405457, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0227, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12752027809619904, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0241, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.11534811556339264, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0234, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.1467696875333786, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0245, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10596702992916107, + "learning_rate": 2.558641420481684e-05, + "loss": 0.02, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.14685864746570587, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.022, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.13554386794567108, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.026, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.12572740018367767, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0232, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.1583555042743683, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0209, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.12913484871387482, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0226, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.11418721824884415, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0225, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.12722279131412506, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.0211, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.1382816582918167, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.0243, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.11612103134393692, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0212, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.1537163406610489, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.021, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.10462892800569534, + "learning_rate": 2.555415209239832e-05, + "loss": 0.0214, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.12759479880332947, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0232, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.1290607452392578, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0222, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12458004057407379, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0213, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.1113182008266449, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0223, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.15024979412555695, + "learning_rate": 2.553945541728904e-05, + "loss": 0.022, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.12657828629016876, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.025, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.13719463348388672, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.024, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.13465221226215363, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0254, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.08891251683235168, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0215, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.08946283161640167, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0224, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.13241387903690338, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0229, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.13068708777427673, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0215, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.12115749716758728, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.025, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.12863430380821228, + "learning_rate": 2.551295099554365e-05, + "loss": 0.0202, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.1332341581583023, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0204, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.102240651845932, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0199, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.16963273286819458, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0226, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12840960919857025, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0225, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.09958021342754364, + "learning_rate": 2.549819836273345e-05, + "loss": 0.02, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.10737278312444687, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0226, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.11473958194255829, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0219, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.12188851088285446, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0216, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.1229034960269928, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0227, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.1316317766904831, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1230575442314148, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0218, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13473588228225708, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0221, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1293632537126541, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0223, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.11663107573986053, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0198, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.147003173828125, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0216, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.10070289671421051, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0204, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12855689227581024, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0222, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.10310211777687073, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0227, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.25780248641967773, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0204, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.11621664464473724, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0196, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.15240156650543213, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0232, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.1359490305185318, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0191, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.13072653114795685, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.0242, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.1296282559633255, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0214, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.12183726578950882, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0215, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.10602327436208725, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0222, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.12424091249704361, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0215, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.14287705719470978, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0217, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.10017845034599304, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0214, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.1265304833650589, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0214, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.15279516577720642, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0194, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.11116814613342285, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0197, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.14380785822868347, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0187, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.12522602081298828, + "learning_rate": 2.541224092500406e-05, + "loss": 0.023, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.12386767566204071, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.0197, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11278095096349716, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0205, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12119927257299423, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0206, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.14128951728343964, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0197, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.12576138973236084, + "learning_rate": 2.539735331078111e-05, + "loss": 0.024, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.13276934623718262, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0217, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.15223100781440735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0222, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.13918957114219666, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0211, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.10129881650209427, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0192, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.15048135817050934, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0191, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.11579657346010208, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.019, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.1882474422454834, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0241, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.12608706951141357, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.021, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.14131402969360352, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0234, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.11546603590250015, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0208, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.0989922359585762, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0219, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.13154740631580353, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0234, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.12450987845659256, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0219, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.15696841478347778, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0228, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12059786915779114, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0236, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.1058221235871315, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0234, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.11429761350154877, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0232, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.11841994524002075, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0193, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.12387857586145401, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0239, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.11453676968812943, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0224, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.15571972727775574, + "learning_rate": 2.533461009735874e-05, + "loss": 0.022, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.10241609811782837, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.022, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.11682361364364624, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0211, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.13623486459255219, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0215, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.12484022229909897, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0212, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.09105533361434937, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0223, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.1024160087108612, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0174, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.14108790457248688, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0215, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.1268249899148941, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0211, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12255556881427765, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0192, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.15904900431632996, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0259, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.1356613039970398, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.023, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.12410393357276917, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0223, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1291753053665161, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0221, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.109084852039814, + "learning_rate": 2.529258887162501e-05, + "loss": 0.0243, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.16595998406410217, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0201, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.14062269032001495, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0232, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.13198593258857727, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0214, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.10668856650590897, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0229, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.11266572028398514, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0233, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.12742944061756134, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0221, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.12420892715454102, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0215, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.1265600174665451, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0215, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.11914686858654022, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0222, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.12876294553279877, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0233, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.09800750762224197, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0219, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1273864060640335, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0238, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.13299106061458588, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0226, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.09396592527627945, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.12955208122730255, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0204, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.16461864113807678, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0211, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.11535970121622086, + "learning_rate": 2.524135719213194e-05, + "loss": 0.02, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.11860322207212448, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0212, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10968036204576492, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0207, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.12494095414876938, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.02, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.10398012399673462, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0205, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.12939494848251343, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0216, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.10996326059103012, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0199, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.13704033195972443, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0208, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.11454308778047562, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.0212, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.14186057448387146, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0226, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.12624409794807434, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0199, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.14392296969890594, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.15774424374103546, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0213, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.1352480947971344, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0192, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.10594955086708069, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0231, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.11280091851949692, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0202, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11683343350887299, + "learning_rate": 2.519293379145164e-05, + "loss": 0.022, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.11721377819776535, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0223, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.18298691511154175, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.0225, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12927964329719543, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0202, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.1413726508617401, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.021, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.1330733746290207, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0228, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12341104447841644, + "learning_rate": 2.517472387591168e-05, + "loss": 0.021, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.12199880182743073, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.0202, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.12259527295827866, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.0227, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.11682213097810745, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0246, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.12783272564411163, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0204, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.09663784503936768, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0228, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.13686567544937134, + "learning_rate": 2.515648614306516e-05, + "loss": 0.02, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.12164216488599777, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0234, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.1371183544397354, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0217, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.12031334638595581, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0203, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.1183433011174202, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.1265946477651596, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0224, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.12686678767204285, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0202, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.13161398470401764, + "learning_rate": 2.513517369661726e-05, + "loss": 0.0216, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.13780610263347626, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0205, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.14213374257087708, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0237, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.12532930076122284, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0215, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1276065707206726, + "learning_rate": 2.512297821731217e-05, + "loss": 0.02, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.08893107622861862, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12819649279117584, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0202, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.09707193821668625, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0202, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.511077043763402e-05, + "loss": 0.0208, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.10047636926174164, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0219, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.15637417137622833, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0218, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.15387912094593048, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0213, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.14876650273799896, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0222, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.13631325960159302, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0214, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.12577500939369202, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0232, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15896311402320862, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0237, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.16459719836711884, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0223, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11081109195947647, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0223, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.12327340245246887, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0219, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14812195301055908, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0228, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.11862681806087494, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0252, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.10684223473072052, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0227, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.1279568374156952, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0216, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.13341103494167328, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0213, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.11913862824440002, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0214, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.13493330776691437, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.0214, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.1272144317626953, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0209, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.11348695307970047, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0217, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.16526950895786285, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0198, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.14953579008579254, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0231, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.12980960309505463, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0226, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.16652929782867432, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0244, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1722235530614853, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.026, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.23629626631736755, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0221, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.12417787313461304, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.023, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.11768155544996262, + "learning_rate": 2.502804732645032e-05, + "loss": 0.0186, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.13731959462165833, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.023, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.18034201860427856, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.0201, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11240512877702713, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0217, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.10463474690914154, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0201, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.1273874044418335, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0194, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.12314468622207642, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0237, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.13170552253723145, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.021, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.12725704908370972, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0198, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.11761748790740967, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0198, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13019445538520813, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0237, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.09089519083499908, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.0202, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.13669925928115845, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0236, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.0891583263874054, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0189, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.12015428394079208, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.0214, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.12889859080314636, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0214, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.1092396154999733, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.022, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.1011805608868599, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0201, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.11652060598134995, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.019, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.12280789762735367, + "learning_rate": 2.496950169240971e-05, + "loss": 0.0229, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.1253744661808014, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0245, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.15854160487651825, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0257, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.11671199649572372, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0204, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.15195102989673615, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.11878954619169235, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0207, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.1035882979631424, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0195, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12105780094861984, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0191, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.13278989493846893, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0207, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.16576215624809265, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.0228, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.13104894757270813, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0197, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.14476071298122406, + "learning_rate": 2.49354818242132e-05, + "loss": 0.023, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.14116041362285614, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0204, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.133507639169693, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0213, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11147522926330566, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0192, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.1317354440689087, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0229, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.15885432064533234, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0214, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.0996033102273941, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0213, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.11059842258691788, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0181, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.1103382259607315, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0252, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.12147006392478943, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.1167205348610878, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0217, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.11838674545288086, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0223, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.1745499074459076, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0213, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.09892687201499939, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0213, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.17135944962501526, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0228, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.13930797576904297, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0211, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.10764696449041367, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0204, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11211943626403809, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0207, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1359340399503708, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0215, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.11819283664226532, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0213, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.1166793555021286, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0216, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.14411170780658722, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0206, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.1380920261144638, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0211, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.12901239097118378, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.0224, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.1171313151717186, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.0207, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.15169396996498108, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0225, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.2066943198442459, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0182, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1179543286561966, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0226, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.12273375689983368, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.0224, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.09459412097930908, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0177, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.10208278894424438, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0207, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.11947600543498993, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.0246, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10635806620121002, + "learning_rate": 2.483599695058918e-05, + "loss": 0.021, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.15165211260318756, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0242, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.10573819279670715, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0221, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.10153646767139435, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0212, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.12596352398395538, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.021, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.11259406059980392, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0214, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.09709548950195312, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0186, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1462256908416748, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0228, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.15480898320674896, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0245, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.10707002878189087, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0197, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.10963082313537598, + "learning_rate": 2.480475069778109e-05, + "loss": 0.019, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.13645771145820618, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0229, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.08208916336297989, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0194, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.11223483830690384, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0212, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.12233787775039673, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0228, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.12175288051366806, + "learning_rate": 2.478909963381241e-05, + "loss": 0.0204, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.14267505705356598, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0203, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.12649142742156982, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0217, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.11254285275936127, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0214, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.1454164832830429, + "learning_rate": 2.477656539969673e-05, + "loss": 0.0221, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.17177407443523407, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0212, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.1089746281504631, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0222, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.1525813490152359, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0236, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.09651731699705124, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0214, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.12710145115852356, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0233, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.15999987721443176, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0215, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1600969433784485, + "learning_rate": 2.475460191424065e-05, + "loss": 0.0201, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.11045356094837189, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0208, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.14192573726177216, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.0216, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.11450321227312088, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0202, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.13454747200012207, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.021, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.11518846452236176, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0194, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.10673099011182785, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0214, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.12131883949041367, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0203, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.18098120391368866, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0182, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.09981405735015869, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0214, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.09194957464933395, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0209, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.11133445054292679, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0227, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.14570456743240356, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0181, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.14878204464912415, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0213, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.1475936472415924, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0229, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.14091694355010986, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0229, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.10411489754915237, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.021, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.10015816241502762, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0211, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09339930862188339, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0206, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.13016946613788605, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0223, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.10682770609855652, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0258, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.13658446073532104, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0232, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.12742926180362701, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.13928014039993286, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0208, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.10824346542358398, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0199, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.11346063017845154, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.0225, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12237603217363358, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0226, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.12201637774705887, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0209, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.09991955012083054, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0239, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.15387621521949768, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.025, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.1588744819164276, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0192, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.1347019523382187, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0196, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.12746664881706238, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0212, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11522521078586578, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0197, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.09697534143924713, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0219, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12183485180139542, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0195, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.13119301199913025, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0195, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.10748504102230072, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0192, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.1617487072944641, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0224, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.11518653482198715, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0207, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.17646238207817078, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0233, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.13729365170001984, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.09777814894914627, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0207, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.117672860622406, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0254, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.09014946967363358, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0214, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.1273731142282486, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0223, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10932891070842743, + "learning_rate": 2.460937119575347e-05, + "loss": 0.0224, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.1672651767730713, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0205, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.0930962786078453, + "learning_rate": 2.46030216185615e-05, + "loss": 0.02, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.09708375483751297, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0192, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.13240690529346466, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0214, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.10310725122690201, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0204, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.13352994620800018, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0206, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.15277482569217682, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0222, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.14223086833953857, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0229, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12845414876937866, + "learning_rate": 2.458077513854331e-05, + "loss": 0.0221, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13195961713790894, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0232, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.1461530476808548, + "learning_rate": 2.457441244940383e-05, + "loss": 0.022, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13729079067707062, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0196, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.15615087747573853, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0213, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.12907880544662476, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0198, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.11320316791534424, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0225, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.14982816576957703, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0225, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.12824060022830963, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0196, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.11940523982048035, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0204, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946357995271683, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.0218, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.13296495378017426, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0229, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.1383914053440094, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0207, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.11321834474802017, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0202, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.1276783049106598, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0193, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1406320035457611, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0191, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.12060298770666122, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0195, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.08627335727214813, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0191, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1429424285888672, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0197, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.15133187174797058, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0212, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.11892743408679962, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0231, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.11871744692325592, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0211, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.105750173330307, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.0234, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10969540476799011, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.022, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.09607745707035065, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0173, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.11796857416629791, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0193, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.114372618496418, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.02, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.1164821982383728, + "learning_rate": 2.449463406055895e-05, + "loss": 0.023, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.13561253249645233, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.018, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.14956215023994446, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0245, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.12785622477531433, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0222, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.16085495054721832, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.021, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10792174935340881, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0209, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.10508404672145844, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0217, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.11471311002969742, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0202, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.14265185594558716, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0224, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.13569262623786926, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0196, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15635624527931213, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0205, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12042741477489471, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0179, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.09488346427679062, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0198, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.20503900945186615, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0199, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.13667535781860352, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13254515826702118, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0204, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.13879507780075073, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0233, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.11448678374290466, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0237, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1255042403936386, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0204, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11896912008523941, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0207, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.12624827027320862, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0194, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.09312308579683304, + "learning_rate": 2.442727178914008e-05, + "loss": 0.022, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12632645666599274, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0189, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.11468527466058731, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0182, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.09245705604553223, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0188, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.12113099545240402, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0191, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.13049040734767914, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0221, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.16979938745498657, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0211, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.11175577342510223, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0172, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.1191241443157196, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0206, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.16317729651927948, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.0213, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.10087717324495316, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0194, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.12195733189582825, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0253, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11880146712064743, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0222, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.14213789999485016, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0198, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.15624478459358215, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.0205, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09695806354284286, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0219, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11260177940130234, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0206, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.15394112467765808, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0219, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.11663617193698883, + "learning_rate": 2.436928131964383e-05, + "loss": 0.0214, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.12596042454242706, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0225, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.10310690104961395, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0203, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.12332364916801453, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0203, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.16257967054843903, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0234, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.13102053105831146, + "learning_rate": 2.435313188161029e-05, + "loss": 0.022, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.14736537635326385, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0199, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.14611469209194183, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0238, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.14594200253486633, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0212, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.16921328008174896, + "learning_rate": 2.434019954415699e-05, + "loss": 0.0236, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.1318638175725937, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0201, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.10869060456752777, + "learning_rate": 2.433372911848899e-05, + "loss": 0.019, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.1446760594844818, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.021, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.1298162341117859, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10307328402996063, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0197, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.11932921409606934, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0209, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.10986137390136719, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0227, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.12099824845790863, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0182, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.13432320952415466, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.0213, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.14018790423870087, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0199, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.13015322387218475, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0191, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.10281360149383545, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0215, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.09500012546777725, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0216, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.15131708979606628, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0225, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.08750247210264206, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0221, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.10143480449914932, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0214, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.12935201823711395, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0202, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.10703578591346741, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0207, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08649012446403503, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0184, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.11186458170413971, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.021, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.11181459575891495, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0211, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.11756999045610428, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0216, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.12061341106891632, + "learning_rate": 2.426561884159603e-05, + "loss": 0.019, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.13861820101737976, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0214, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.11545911431312561, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0195, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.14892302453517914, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0226, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.10354511439800262, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0197, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.15290679037570953, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0208, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.12185964733362198, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0205, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.10486926138401031, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0206, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.10959216207265854, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0198, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.11638635396957397, + "learning_rate": 2.423633359536448e-05, + "loss": 0.0227, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.13661672174930573, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0206, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.13071168959140778, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0189, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.16377446055412292, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0203, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.22221329808235168, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0185, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.11251460015773773, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0191, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.10873523354530334, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0228, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.13705629110336304, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0203, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.13980528712272644, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.10355913639068604, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0199, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.1282302737236023, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.0215, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.1221374049782753, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0192, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.18542620539665222, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0199, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1250954121351242, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0189, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12665164470672607, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0217, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.09464617818593979, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.0197, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.09657798707485199, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0194, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.11533467471599579, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0189, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.13647402822971344, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0194, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.13276426494121552, + "learning_rate": 2.41743228236291e-05, + "loss": 0.022, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.16899874806404114, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0209, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.14313359558582306, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0211, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.14507170021533966, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0218, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.13025349378585815, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0209, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.1666184365749359, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0217, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14891882240772247, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0205, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11392978578805923, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0223, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.12964752316474915, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.019, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11125730723142624, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0211, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.08378217369318008, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0203, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.14530996978282928, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0206, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.10187243670225143, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0188, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.09652446955442429, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0192, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.10552513599395752, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0212, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11044801771640778, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0215, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.12248426675796509, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0185, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.11164992302656174, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0198, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.11919678002595901, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0187, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1675145924091339, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0262, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.15451745688915253, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0226, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.0997307077050209, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0196, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.1272345781326294, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0214, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.1286081224679947, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0214, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12226564437150955, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0211, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.11795010417699814, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0216, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.12963736057281494, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0215, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.10415369272232056, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0215, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.11642655730247498, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0194, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.19006073474884033, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0216, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.16677096486091614, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.0235, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.15928123891353607, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0195, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.10270075500011444, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0197, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.12253808975219727, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0216, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14145877957344055, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0211, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1387409120798111, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0185, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.11433693021535873, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0204, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.14247572422027588, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0187, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.12203793972730637, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.018, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.15705470740795135, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0212, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.10648002475500107, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0182, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.11573450267314911, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0211, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.11960671842098236, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0245, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.14416365325450897, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0178, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.11234421283006668, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0205, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.1375260204076767, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.025, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.08686859905719757, + "learning_rate": 2.402315322351119e-05, + "loss": 0.018, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11352542787790298, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0198, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.13374295830726624, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0216, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.09358088672161102, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0202, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.13767337799072266, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0229, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.1405765861272812, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.02, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11317892372608185, + "learning_rate": 2.400332803040664e-05, + "loss": 0.02, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.12356667965650558, + "learning_rate": 2.400002143698994e-05, + "loss": 0.0225, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.12570610642433167, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0232, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.1461946964263916, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0201, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.10893318802118301, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0211, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.10969080775976181, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0235, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.17320629954338074, + "learning_rate": 2.398347822254704e-05, + "loss": 0.0221, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14296430349349976, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0206, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12349261343479156, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0182, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.11617975682020187, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0209, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.1574973613023758, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0222, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.157145157456398, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.0199, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.14698748290538788, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0203, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.12345635890960693, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0212, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.1370248794555664, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0194, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.1332298070192337, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0219, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.12525111436843872, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.0219, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.14905311167240143, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0211, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12063094973564148, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.0262, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.10361988842487335, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0195, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.09919667989015579, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0201, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.11141914129257202, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.0223, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.17785966396331787, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0201, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.126350998878479, + "learning_rate": 2.392710390356112e-05, + "loss": 0.0201, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.12107345461845398, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0213, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.11466097086668015, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0202, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.15397872030735016, + "learning_rate": 2.391713512040237e-05, + "loss": 0.022, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.14412778615951538, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0223, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.10495682060718536, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0194, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.13021183013916016, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0199, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.13993357121944427, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0182, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.128571018576622, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.0181, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.11090676486492157, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0204, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.10988310724496841, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0203, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.11705116927623749, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0214, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.1277252584695816, + "learning_rate": 2.388719222953459e-05, + "loss": 0.0225, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1561414748430252, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0221, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.15278629958629608, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.02, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14865094423294067, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0213, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.1127776950597763, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0211, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.15085943043231964, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0212, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14555132389068604, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0216, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.14708220958709717, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0209, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.1093590185046196, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0199, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.1303861141204834, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0242, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.11462346464395523, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0197, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.12198365479707718, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0206, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.11193065345287323, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0201, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.13254034519195557, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0181, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.13355755805969238, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0204, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12551282346248627, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0212, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1441735178232193, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0217, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.11511780321598053, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0218, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.10427968204021454, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0222, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.11364301294088364, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0166, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.1401003897190094, + "learning_rate": 2.38204570134882e-05, + "loss": 0.0215, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.11645112931728363, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0198, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1340353935956955, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0212, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.0982050895690918, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0208, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.16713382303714752, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0185, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.13803045451641083, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.021, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.15552546083927155, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0233, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.10836438834667206, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0211, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.1831241250038147, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0198, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.12369874119758606, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0193, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.10911241918802261, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0221, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.11441849172115326, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0236, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.11105310916900635, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0206, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.13617953658103943, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0189, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.11214766651391983, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0192, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.12536488473415375, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0206, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.10356712341308594, + "learning_rate": 2.376687582517292e-05, + "loss": 0.019, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.13332195580005646, + "learning_rate": 2.376352133090187e-05, + "loss": 0.0216, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.13996818661689758, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0216, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.14305290579795837, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0204, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.10494769364595413, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0202, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.1461218297481537, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.0197, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.12469783425331116, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0207, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.13164709508419037, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.0207, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.1553804576396942, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0228, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1445176750421524, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0217, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12673451006412506, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0164, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.17502479255199432, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0208, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.11328406631946564, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0196, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12234057486057281, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0193, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.1333971917629242, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0214, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.11227905005216599, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0183, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.1287100911140442, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0198, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.10215240716934204, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0194, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.09786848723888397, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0202, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.11103200167417526, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0219, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.1491294801235199, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0202, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.1364760547876358, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.0198, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.11583423614501953, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0206, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.13896186649799347, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0183, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.0970211774110794, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.019, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.10326618701219559, + "learning_rate": 2.368281439559657e-05, + "loss": 0.022, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.11581890285015106, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0234, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10186298936605453, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0205, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11655393242835999, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0193, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.11990638077259064, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0217, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.10573728382587433, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0177, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.09257248789072037, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.0216, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.11310117691755295, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0188, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.12632012367248535, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0217, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1535339206457138, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0209, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.13133156299591064, + "learning_rate": 2.364907420469738e-05, + "loss": 0.0207, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11225332319736481, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0217, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.10744500905275345, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0212, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.10509403049945831, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0208, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.16215650737285614, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0199, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1208849623799324, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0196, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.14882729947566986, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0229, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.18775896728038788, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0206, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.14199234545230865, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0201, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.1510499119758606, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.021, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.16389605402946472, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0197, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.13032132387161255, + "learning_rate": 2.361188413960337e-05, + "loss": 0.0211, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.13895465433597565, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0222, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.10914628207683563, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0206, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.12215261161327362, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0186, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12954379618167877, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.0193, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.12779726088047028, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.02, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.11825352907180786, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0214, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11001764982938766, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0211, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.10464788973331451, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.02, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.0892137810587883, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0198, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.09811265021562576, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0191, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.14199955761432648, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0196, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.13846102356910706, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0216, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.12102426588535309, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0214, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.12360251694917679, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0199, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.12587665021419525, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.15562282502651215, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.0214, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.11366713047027588, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0192, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.15177540481090546, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0204, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.08908293396234512, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0186, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.1349143236875534, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0197, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.09388674795627594, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0205, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.11973906308412552, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0198, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.1290522664785385, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.02, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.12087979167699814, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0207, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.10237318277359009, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0189, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.10426577180624008, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0204, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13028934597969055, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0197, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.10696206241846085, + "learning_rate": 2.351686222335574e-05, + "loss": 0.0206, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.11024264991283417, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0189, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.11294405907392502, + "learning_rate": 2.35100554725415e-05, + "loss": 0.0228, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.1188240721821785, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0204, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.17485342919826508, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0202, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12582673132419586, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0209, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.18180538713932037, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0202, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.1459868997335434, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0189, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.09818759560585022, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0195, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.1473611295223236, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0253, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.09137327969074249, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0209, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.14026956260204315, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0202, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.1511334776878357, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0182, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.15016451478004456, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0204, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.14407901465892792, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.0205, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.1332983374595642, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0192, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.14312349259853363, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.0195, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.11917808651924133, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0194, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.17896629869937897, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0203, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.12084568291902542, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0199, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.14310620725154877, + "learning_rate": 2.344867863274048e-05, + "loss": 0.0201, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1168353408575058, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0225, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.14718633890151978, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0204, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.14102597534656525, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0197, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.09543738514184952, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0218, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1074792668223381, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0198, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.11555791646242142, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0196, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.12645067274570465, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0191, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.11634501069784164, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.02, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.09201087802648544, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0204, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.1155405044555664, + "learning_rate": 2.341449046471072e-05, + "loss": 0.0215, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.09871169179677963, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0216, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.11332481354475021, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0179, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.12474322319030762, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.0177, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.11339092999696732, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0221, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.1200975552201271, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.0209, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.14456698298454285, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0211, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.1371302306652069, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0165, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.11727641522884369, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0232, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.11336934566497803, + "learning_rate": 2.338366646946928e-05, + "loss": 0.0202, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.10325363278388977, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0209, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.11268003284931183, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0187, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10260385274887085, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.0195, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.10920321941375732, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0225, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.09120934456586838, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0194, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11567433178424835, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0197, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.10144805908203125, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0206, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1246950775384903, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0215, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.100107841193676, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0197, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.15822021663188934, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0202, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13518433272838593, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.02, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12268074601888657, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0196, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.12333078682422638, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0184, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.13588106632232666, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0218, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.12120439857244492, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0197, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.16428445279598236, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0198, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.11615289747714996, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.0182, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.1432264745235443, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0207, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.11717039346694946, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0198, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.10115417838096619, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0193, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.11894490569829941, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0198, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.11940047889947891, + "learning_rate": 2.330810220141532e-05, + "loss": 0.021, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.10992413759231567, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0195, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.10544339567422867, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0182, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.11725642532110214, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.0198, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.09130900353193283, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0209, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.09033828228712082, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0177, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.11721840500831604, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0207, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.12251118570566177, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0215, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.1496249884366989, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0219, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.14808881282806396, + "learning_rate": 2.327710137160376e-05, + "loss": 0.0212, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.145681232213974, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.019, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.12777844071388245, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0173, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.1110587939620018, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0198, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.11348012834787369, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0207, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.11685867607593536, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.02, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.17935989797115326, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0211, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.12398742884397507, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0191, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.10836582630872726, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0196, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.12038073688745499, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0207, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.12211567908525467, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0206, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.1702142357826233, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0227, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12037540227174759, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0191, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.15964926779270172, + "learning_rate": 2.323223255665661e-05, + "loss": 0.018, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.12493535876274109, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0212, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.1359032839536667, + "learning_rate": 2.32253202698687e-05, + "loss": 0.0219, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.18753379583358765, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0228, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.1336824893951416, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0201, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.09901099652051926, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0221, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.12925978004932404, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0206, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.1278320550918579, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.0178, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.12977559864521027, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0176, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10098735243082047, + "learning_rate": 2.320110760314603e-05, + "loss": 0.023, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.12594632804393768, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.0209, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.1140967383980751, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0181, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.1287144273519516, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0195, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.12329071760177612, + "learning_rate": 2.318725808577064e-05, + "loss": 0.0198, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.1339070051908493, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10581240057945251, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0265, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.13716311752796173, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0213, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.14610108733177185, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0195, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.128855898976326, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0185, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.0950121283531189, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0223, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.14444753527641296, + "learning_rate": 2.316299750241968e-05, + "loss": 0.017, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.12585613131523132, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0177, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.1290416419506073, + "learning_rate": 2.315606032389525e-05, + "loss": 0.02, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.10740494728088379, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0191, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11578185111284256, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0211, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.11840540170669556, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0209, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.09531640261411667, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0202, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.10436122864484787, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0188, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.14102807641029358, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0205, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14070506393909454, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0186, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.09883972257375717, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0196, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11524099856615067, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0187, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.11240902543067932, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0176, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.08062270283699036, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.019, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.13965697586536407, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0202, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.11376950889825821, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0218, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.13540907204151154, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0188, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.10888075083494186, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0216, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.10280206799507141, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0178, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.11512117087841034, + "learning_rate": 2.309699448641602e-05, + "loss": 0.02, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.09106973558664322, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0182, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10531709343194962, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0204, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.159554585814476, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0173, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.12161346524953842, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0196, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.08224594593048096, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0175, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.12767191231250763, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0234, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.12864384055137634, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0224, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.10232329368591309, + "learning_rate": 2.3069137227667e-05, + "loss": 0.0208, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.09609836339950562, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0183, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.1301981806755066, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0202, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.1258714497089386, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0217, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.12543867528438568, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13326771557331085, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.022, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.11770334839820862, + "learning_rate": 2.304821853869955e-05, + "loss": 0.019, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.11729256063699722, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0187, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.11887779086828232, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0224, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.10921293497085571, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0192, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.1346455216407776, + "learning_rate": 2.303426051834282e-05, + "loss": 0.022, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.1177918091416359, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0209, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.12912803888320923, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0191, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.12571996450424194, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0211, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.13959148526191711, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0198, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.1404334306716919, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0211, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.1111592948436737, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0213, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.10322082787752151, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0203, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.11235317587852478, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0187, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.13599546253681183, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0205, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11067865788936615, + "learning_rate": 2.299932279415038e-05, + "loss": 0.02, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.14130094647407532, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0204, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.14370687305927277, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0199, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.13825304806232452, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0197, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.11885961890220642, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.02, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.12376029789447784, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0199, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.10103051364421844, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0181, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.10931621491909027, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0198, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.11835317313671112, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0202, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.08036374300718307, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0189, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.08135218918323517, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0196, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.12611854076385498, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0211, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.09217566251754761, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0181, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.10258878767490387, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0194, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.1165921688079834, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0196, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.09290144592523575, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.0205, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.11575523763895035, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0204, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13866478204727173, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0178, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.13036414980888367, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0191, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11062757670879364, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0193, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.09973529726266861, + "learning_rate": 2.292926536293724e-05, + "loss": 0.02, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.13707594573497772, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0186, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11635670065879822, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0185, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.13926011323928833, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0192, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.13966591656208038, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0194, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.09878037869930267, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0191, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.11935862898826599, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0193, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.15291833877563477, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0189, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.12976837158203125, + "learning_rate": 2.290117482755701e-05, + "loss": 0.017, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.13367557525634766, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0178, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12121888250112534, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0191, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.13135644793510437, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0226, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.1142699271440506, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.017, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.09670589119195938, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0199, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.12751951813697815, + "learning_rate": 2.288008171679078e-05, + "loss": 0.018, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.14463315904140472, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0184, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.11006469279527664, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0198, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.15361396968364716, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0186, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.20782725512981415, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0208, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.15324300527572632, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.018, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.17235179245471954, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0218, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.12993113696575165, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0198, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.1496446281671524, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0203, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.16271376609802246, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.0215, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11106697469949722, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0206, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.10508760809898376, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0218, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.11433881521224976, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0203, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.12877102196216583, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0203, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.13301214575767517, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0225, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.15351711213588715, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0213, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.1114102303981781, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0212, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.09978143125772476, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0198, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10524465143680573, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0183, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.09780002385377884, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0185, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.1361900269985199, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0218, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.1329333484172821, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0191, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.10501913726329803, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.019, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.1329827904701233, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0198, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.10849577933549881, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0213, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.14308017492294312, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.0199, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.12313494831323624, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0193, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.10307908058166504, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.0215, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10591195523738861, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0178, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.1411430686712265, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0237, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13116225600242615, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0203, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.12032114714384079, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0201, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.12519164383411407, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0177, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.10340812057256699, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0194, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.10876207798719406, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0207, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11245796829462051, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0207, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11444602906703949, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0171, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12232744693756104, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0196, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.10049490630626678, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0179, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.09973384439945221, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.022, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.11037951707839966, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0197, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.14473563432693481, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0207, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.1138916090130806, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0198, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.12471994757652283, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0182, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.12418543547391891, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0195, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.12200118601322174, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.0217, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.13389603793621063, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0194, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10176926106214523, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.0189, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.10195910185575485, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0202, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.13289214670658112, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.018, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.10307228565216064, + "learning_rate": 2.270347347405331e-05, + "loss": 0.0183, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.149918794631958, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0172, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.1182866096496582, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0173, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.1035924181342125, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0194, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.108797088265419, + "learning_rate": 2.268928118647676e-05, + "loss": 0.0174, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.11457139253616333, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0208, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.10256518423557281, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0188, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13053470849990845, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0209, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.11632310599088669, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0187, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.12769649922847748, + "learning_rate": 2.267152769015067e-05, + "loss": 0.0206, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.13292407989501953, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0191, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.09822595119476318, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0189, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.13729946315288544, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0229, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.15705017745494843, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0203, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.12355367839336395, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0201, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.1278308629989624, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0205, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.09874208271503448, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0218, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.1316995620727539, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0212, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.10361800342798233, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0209, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.09672107547521591, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0205, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.12609392404556274, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0205, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.1360992044210434, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0188, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.11627763509750366, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0206, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.11733578145503998, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.02, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.1705910563468933, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0174, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.11704754829406738, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0199, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.13247141242027283, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0224, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.10916013270616531, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0202, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.11092681437730789, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.0186, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.12811405956745148, + "learning_rate": 2.260036839187291e-05, + "loss": 0.02, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.11893392354249954, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0194, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.16553092002868652, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0189, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.12431131303310394, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.02, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10850611329078674, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0222, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.12647055089473724, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0193, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.10432784259319305, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0196, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.1159503236413002, + "learning_rate": 2.257540798155951e-05, + "loss": 0.0192, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.10592516511678696, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0219, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.1318269520998001, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0195, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.11932560801506042, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0194, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.16499531269073486, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.02, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.1334158480167389, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.0217, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.14781776070594788, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0206, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.14816460013389587, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0206, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.11319883167743683, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0196, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.12413972616195679, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0206, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.1277540624141693, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0211, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.12178509682416916, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0192, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.1210661232471466, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0232, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.11300696432590485, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0182, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.15876832604408264, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0194, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.1387646645307541, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0202, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.12502169609069824, + "learning_rate": 2.251824990920905e-05, + "loss": 0.02, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.12365051358938217, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0197, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.12610606849193573, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0194, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.13096576929092407, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0208, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.13002553582191467, + "learning_rate": 2.250393750922939e-05, + "loss": 0.02, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.10048994421958923, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0193, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.11463778465986252, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0179, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10637608915567398, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0182, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.11959703266620636, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0183, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.11852603405714035, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12075617164373398, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0186, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.09626025706529617, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0195, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12773670256137848, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.019, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13364142179489136, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.019, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.14562146365642548, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0207, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.11752697080373764, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.13970085978507996, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0219, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.12456551194190979, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0203, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.36701101064682007, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0191, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.17254851758480072, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0193, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.1397981345653534, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0205, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.1273665875196457, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0227, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.08652572333812714, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0194, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.118954136967659, + "learning_rate": 2.243582938893974e-05, + "loss": 0.019, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.12105695903301239, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0173, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.10059279948472977, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0189, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.1476186215877533, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0192, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.11198431998491287, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0212, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.14829197525978088, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0216, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.14479608833789825, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0225, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.16737256944179535, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0202, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.12101127207279205, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0198, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.09408389776945114, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.019, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.10649813711643219, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0191, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.1328127533197403, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.02, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1135721281170845, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0181, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.14464536309242249, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0167, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.15146523714065552, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0216, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11212945729494095, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0189, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.11722676455974579, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0185, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10031717270612717, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0181, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.14852827787399292, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.02, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12114712595939636, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0207, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.12325385212898254, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0185, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.12915199995040894, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0188, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.10423776507377625, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0195, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.13398106396198273, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0182, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.1426725685596466, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0214, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.12515439093112946, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0184, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.15888792276382446, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0183, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.11037078499794006, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0199, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.10706029832363129, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.0195, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.11116144061088562, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0202, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.12716875970363617, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0182, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.1446460634469986, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0172, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.09947994351387024, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0205, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.09251125156879425, + "learning_rate": 2.231705291760663e-05, + "loss": 0.018, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.08440345525741577, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0181, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12813445925712585, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0206, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.1017315611243248, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0192, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.13247326016426086, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0212, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.14845387637615204, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0194, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.13824601471424103, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0224, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.1151830404996872, + "learning_rate": 2.229177978101565e-05, + "loss": 0.0176, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.12420724332332611, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0179, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.13085328042507172, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0179, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.12436439841985703, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.11873050034046173, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0197, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.1338828206062317, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0212, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.1106492280960083, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0179, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.16542823612689972, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0202, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.08829429000616074, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0176, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.10792568325996399, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0186, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.1279839128255844, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0221, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.10742732882499695, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0217, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.10220900177955627, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0201, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.1660427749156952, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.019, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.13747647404670715, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0225, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.13351675868034363, + "learning_rate": 2.223753180973339e-05, + "loss": 0.021, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.14151513576507568, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.0195, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.12823058664798737, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0189, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.12779751420021057, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0177, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.2243196815252304, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0183, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.12365520745515823, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0188, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.10981941968202591, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0191, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.11569637805223465, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0195, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.15785852074623108, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0222, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11610189825296402, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0172, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.1124223843216896, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0184, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.14634257555007935, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.0196, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.13306891918182373, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0204, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.08198657631874084, + "learning_rate": 2.219041683306619e-05, + "loss": 0.0177, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.09529033303260803, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0166, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.11170824617147446, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0195, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13825464248657227, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0191, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.11621235311031342, + "learning_rate": 2.217590132288247e-05, + "loss": 0.0202, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11921709030866623, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.019, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12485305964946747, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0205, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.1244436725974083, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0204, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11240405589342117, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0197, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.1087997704744339, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0183, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.11403695493936539, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0172, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.12764202058315277, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0161, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.17418256402015686, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0171, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.1522650271654129, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0197, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.16773882508277893, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0214, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.14681261777877808, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0208, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.14842623472213745, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.021, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.10614728182554245, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0182, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.15053948760032654, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0209, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.1381484866142273, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.0176, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.13161014020442963, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0174, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.09695997089147568, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0202, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.08903522044420242, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0171, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11456362903118134, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0183, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.1260307878255844, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.02, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.11732659488916397, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0197, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.126072496175766, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0196, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.11586430668830872, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.02, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.10504680871963501, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0193, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.09652367234230042, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0186, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10370496660470963, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0176, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.1192321851849556, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0178, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.10527302324771881, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0182, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10162831097841263, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0191, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.16128556430339813, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0203, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.11897512525320053, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0204, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.14880484342575073, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0205, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.11369120329618454, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.0193, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.11618658900260925, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0176, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.10052158683538437, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.0207, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1014244481921196, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0205, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.15118759870529175, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0215, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.09245041012763977, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0176, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.13509149849414825, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0224, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.12747371196746826, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0185, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11445175856351852, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0208, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11949758231639862, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0193, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.12797918915748596, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0182, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0185, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.11956185847520828, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0189, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.11755053699016571, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0184, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.1273607611656189, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0205, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.11855774372816086, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0188, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.09031623601913452, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0183, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12208469212055206, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0199, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.1083436831831932, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0193, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12228160351514816, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0182, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.1007765457034111, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0195, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.11114434152841568, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0173, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.1096438392996788, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0199, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.124867744743824, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0219, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.10032925009727478, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.0194, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.13164062798023224, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0193, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.1376691609621048, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0187, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.11254525184631348, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0186, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.12835443019866943, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.15313562750816345, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0219, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.16989216208457947, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0172, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.11433102935552597, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.019, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09817333519458771, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0222, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.09500890970230103, + "learning_rate": 2.193514973491427e-05, + "loss": 0.0176, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.14314892888069153, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0199, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.11764178425073624, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0169, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.12384455651044846, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0193, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.11438048630952835, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0174, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.13921929895877838, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0182, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.16880372166633606, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0174, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.1554516702890396, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0194, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.15355393290519714, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0212, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.1179167702794075, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0195, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.13388752937316895, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0208, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.10997489839792252, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0181, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.11783649772405624, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0189, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13451549410820007, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0186, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.1339322179555893, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0179, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.09972941130399704, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0214, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.10101358592510223, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0176, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.09771978110074997, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0177, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.14234617352485657, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0184, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.10462982952594757, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0197, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.09772231429815292, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0226, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.12018870562314987, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0164, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.10980863869190216, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0199, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.11285397410392761, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0172, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.1067347601056099, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0183, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.1046559289097786, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.0198, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.11118492484092712, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0179, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.1001342311501503, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0166, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.1053142324090004, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0188, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.16253869235515594, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.018, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.10042452812194824, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0173, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.11815430223941803, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0181, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.07585619390010834, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0195, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.17831926047801971, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0183, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.09953836351633072, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0194, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.12129344046115875, + "learning_rate": 2.18065424687869e-05, + "loss": 0.0188, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.09982063621282578, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0196, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.0906965360045433, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0172, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.1256277859210968, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.019, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.10290617495775223, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0171, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.17007306218147278, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0178, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.0779242068529129, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0192, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.12276264280080795, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0233, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.13211409747600555, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0184, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11407056450843811, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0211, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.13331526517868042, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0193, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.1603303849697113, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0192, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.10634022206068039, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0174, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.1658177524805069, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0207, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.1432579606771469, + "learning_rate": 2.175492159090619e-05, + "loss": 0.02, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.11264102905988693, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0183, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.12988954782485962, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.0199, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.1687532663345337, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0175, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12261682003736496, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0148, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.13096439838409424, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0175, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.13871102035045624, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0198, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.12536168098449707, + "learning_rate": 2.172907339903944e-05, + "loss": 0.018, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.12913931906223297, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0194, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.1255621761083603, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0172, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.12288706004619598, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0206, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.113137386739254, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0201, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.11107201129198074, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.14609375596046448, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.02, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.15227735042572021, + "learning_rate": 2.170320016677318e-05, + "loss": 0.016, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.12586566805839539, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0194, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.13495711982250214, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0187, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.11436781287193298, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.018, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.12271122634410858, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0193, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.16221405565738678, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0195, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.08645932376384735, + "learning_rate": 2.168100325341058e-05, + "loss": 0.0157, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.11750698834657669, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0186, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.1191735714673996, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0178, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.10067173838615417, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0187, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.09761539846658707, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0169, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.13806267082691193, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0217, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.10618548095226288, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0165, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.15135695040225983, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0199, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10563432425260544, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0176, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.12757790088653564, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0183, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.11258719861507416, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.0162, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.15290242433547974, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0192, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.1350955218076706, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0188, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.09494426846504211, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0187, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12816165387630463, + "learning_rate": 2.162913952762737e-05, + "loss": 0.019, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.1237507089972496, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0187, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.12867869436740875, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0202, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11882596462965012, + "learning_rate": 2.161801300808089e-05, + "loss": 0.019, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.10291648656129837, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0187, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.09269602596759796, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.019, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.08865118771791458, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.0188, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10444393008947372, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0182, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.11461203545331955, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.0219, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.13728564977645874, + "learning_rate": 2.159574640654742e-05, + "loss": 0.0223, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.1394069343805313, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0172, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.15394151210784912, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0182, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.10422483086585999, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0175, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.10996434837579727, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0182, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.1026761382818222, + "learning_rate": 2.157717712784381e-05, + "loss": 0.018, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.1373908817768097, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0226, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.10155291855335236, + "learning_rate": 2.156974591793542e-05, + "loss": 0.0179, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.14904215931892395, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0192, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.1131916344165802, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0175, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.09735766798257828, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0182, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.13132493197917938, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0193, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.12934266030788422, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0184, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.1196565330028534, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0186, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.12375814467668533, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.0199, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.0892246812582016, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0178, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.0932748094201088, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0192, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11503782123327255, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0187, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.08250989019870758, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0183, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.101202592253685, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0183, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1226576566696167, + "learning_rate": 2.152139450890464e-05, + "loss": 0.022, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.1167614683508873, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0176, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.12396091967821121, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0186, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.13461633026599884, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.0169, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.10792742669582367, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0189, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11649876832962036, + "learning_rate": 2.150277549265773e-05, + "loss": 0.02, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.1314917951822281, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0157, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.14143578708171844, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0197, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10042720288038254, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0177, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.12019779533147812, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0164, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.14096254110336304, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0223, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.10627058893442154, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0171, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.11092417687177658, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.019, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.12211407721042633, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0199, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11551571637392044, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0197, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.10474767535924911, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.018, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.12364303320646286, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0178, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13243450224399567, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0197, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10594736039638519, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0179, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.11143538355827332, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0196, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.10518378764390945, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0197, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12268701940774918, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0176, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.09408178925514221, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0205, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.12529458105564117, + "learning_rate": 2.143564506261267e-05, + "loss": 0.0184, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.15819092094898224, + "learning_rate": 2.143191093594534e-05, + "loss": 0.021, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.15818199515342712, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0182, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.13374921679496765, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0177, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.10046645253896713, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0197, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.11422070860862732, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0185, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.1138019785284996, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0182, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.09617039561271667, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0178, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.1452113538980484, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0222, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.10409227013587952, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0196, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.11277221143245697, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0175, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.12325838208198547, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0201, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12721778452396393, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0174, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.14016005396842957, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.017, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.11149191111326218, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0168, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.13445520401000977, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0194, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10086208581924438, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0197, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.115168996155262, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0181, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.11483300477266312, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.0183, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.1196938082575798, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0172, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.12219196557998657, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0177, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.11318044364452362, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.0204, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.09219531714916229, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0181, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.09911682456731796, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0192, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.11822745203971863, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0192, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.11101230978965759, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0165, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.09978491812944412, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0203, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.09874484688043594, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0199, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.08682837337255478, + "learning_rate": 2.133090580749213e-05, + "loss": 0.0177, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.14352355897426605, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0189, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.11133426427841187, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0199, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.10667043924331665, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0183, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.14207594096660614, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0218, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.12165568023920059, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0167, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.09311263263225555, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0183, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10423581302165985, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0151, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.12640471756458282, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0214, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.11349574476480484, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0171, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.14359329640865326, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0182, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.10977838933467865, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0181, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.12443531304597855, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0168, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.12441008538007736, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0186, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.14103129506111145, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0197, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10460726171731949, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0178, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.12526848912239075, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0188, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.13363681733608246, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0183, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.13106584548950195, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0181, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.1353096216917038, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0202, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.11889169365167618, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0198, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.12628430128097534, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0178, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.13260240852832794, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0182, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12357809394598007, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0175, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.13913387060165405, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0189, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.12303008884191513, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0187, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.10231908410787582, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.017, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.20065239071846008, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0172, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.11311347782611847, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.019, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.131951242685318, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0182, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.15945149958133698, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0202, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09272884577512741, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0189, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.0937163308262825, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0208, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.10963460803031921, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0197, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.1618727743625641, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0195, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.0919048935174942, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0177, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.12060586363077164, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0209, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.13121892511844635, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0181, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.11795026063919067, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0186, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.13060589134693146, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0226, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.08142203092575073, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0181, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.12634120881557465, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0178, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.1448773890733719, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0179, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1230989471077919, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0188, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.11944684386253357, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0212, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.1445692479610443, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0184, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.0831015482544899, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0171, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.1456967294216156, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0189, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.11207464337348938, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0185, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.16137553751468658, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0182, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.09820980578660965, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0192, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.11823676526546478, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0183, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.12873874604701996, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0181, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.11243417114019394, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0188, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.11120739579200745, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0222, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.10704528540372849, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0178, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.12063440680503845, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0174, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.13275623321533203, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0171, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.10400331765413284, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0201, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.10485411435365677, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0187, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.12517869472503662, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0187, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.11703156679868698, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.0178, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.11982452869415283, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0186, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11858765780925751, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0182, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.10404367744922638, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0173, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.09697552025318146, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0188, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.12046534568071365, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0173, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.0896986722946167, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0188, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.10993069410324097, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0213, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.1200791671872139, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0179, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.14664512872695923, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.0202, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.09285914152860641, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0207, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.1080656424164772, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0188, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.140005961060524, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0188, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.11084222793579102, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.02, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.0913061872124672, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0194, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.09459970891475677, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0182, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.09709502756595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0195, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.11101861298084259, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0184, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.1152782216668129, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0178, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.11495838314294815, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0178, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.14250393211841583, + "learning_rate": 2.10258099183954e-05, + "loss": 0.019, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.09898682683706284, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0173, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.13904057443141937, + "learning_rate": 2.101823807728166e-05, + "loss": 0.018, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.12013959139585495, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0163, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.0925723984837532, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0187, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.1284961700439453, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0158, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.12408998608589172, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0206, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1095876470208168, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0172, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.10166801512241364, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0183, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13473083078861237, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0155, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.10492225736379623, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0188, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.10769215226173401, + "learning_rate": 2.098414220174082e-05, + "loss": 0.021, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.09331739693880081, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0176, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.09247364103794098, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0163, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.20951682329177856, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0203, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.16750548779964447, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0196, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.10933154821395874, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0169, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10767697542905807, + "learning_rate": 2.09613911545763e-05, + "loss": 0.0184, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.11565674096345901, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0208, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.11934554576873779, + "learning_rate": 2.095380384728948e-05, + "loss": 0.018, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.12786556780338287, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0169, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.11752485483884811, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0194, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.11586920917034149, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0179, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.08849998563528061, + "learning_rate": 2.0938623809196e-05, + "loss": 0.018, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.12535157799720764, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0176, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.10429015755653381, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0188, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.09504082053899765, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0186, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.12917572259902954, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0182, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.11084037274122238, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0163, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.12090355157852173, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0196, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.11049258708953857, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0194, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.0831870436668396, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0188, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11006250977516174, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.018, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.12107989192008972, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0178, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.10361241549253464, + "learning_rate": 2.089684155251222e-05, + "loss": 0.017, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.13428105413913727, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0198, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.13375632464885712, + "learning_rate": 2.088923894557553e-05, + "loss": 0.018, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.14561115205287933, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0181, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.09850531816482544, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0176, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.11646582931280136, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0162, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.14659325778484344, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0168, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.15971294045448303, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0198, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.11780509352684021, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0156, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11453750729560852, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0179, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.09094218909740448, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0172, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.13252301514148712, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0209, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.10068004578351974, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0159, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.10302501171827316, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0172, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.10893740504980087, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.019, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.12446500360965729, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0177, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1345142424106598, + "learning_rate": 2.083597073454921e-05, + "loss": 0.021, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.12400489300489426, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0194, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.15042100846767426, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0162, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.1559734046459198, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0189, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.10689649730920792, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0163, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.15904614329338074, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0199, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.11539188027381897, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0208, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.14550678431987762, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0191, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.10212639719247818, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0177, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.10318224877119064, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0176, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0196, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.10269734263420105, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0179, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.11070621758699417, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0196, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13299307227134705, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.02, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11430692672729492, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0188, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10718249529600143, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.0204, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.11855681240558624, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.019, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.10754020512104034, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0157, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11011523753404617, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0172, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.11931152641773224, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.0171, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.12425696104764938, + "learning_rate": 2.075972279799841e-05, + "loss": 0.0195, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.10292454063892365, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0177, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.11290059238672256, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0165, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.1453622728586197, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0207, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.12152253836393356, + "learning_rate": 2.07444521407078e-05, + "loss": 0.017, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.134286567568779, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0198, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.11180262267589569, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0196, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.12045837193727493, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.018, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.18509334325790405, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0209, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.17120368778705597, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0197, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.11996874213218689, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0204, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.12391912192106247, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0201, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.1248779147863388, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.016, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.12873846292495728, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0184, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.11072682589292526, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.0182, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.10367089509963989, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0179, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.12854290008544922, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0184, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.16348853707313538, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0169, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.1345866173505783, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0187, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.1139485239982605, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0178, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.1656051129102707, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0177, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.1133129820227623, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0154, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.0865294560790062, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0163, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.1368715763092041, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0218, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.11601361632347107, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0185, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.1401238590478897, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0177, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.10183753073215485, + "learning_rate": 2.066033925513678e-05, + "loss": 0.0192, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.10317160934209824, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0153, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.12317138910293579, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.021, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.10953490436077118, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0191, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.10530596971511841, + "learning_rate": 2.064502355956822e-05, + "loss": 0.0179, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.09054628014564514, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0189, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.12472135573625565, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0239, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.1551167368888855, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0189, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.08141466975212097, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0174, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.13025878369808197, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0185, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.13056428730487823, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0183, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.1187913790345192, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0182, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.11957527697086334, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0173, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.09339714050292969, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0182, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.1584015190601349, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.02, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.13754191994667053, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0193, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.09880727529525757, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0192, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.15097497403621674, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0172, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12028415501117706, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0182, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.10583297163248062, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0183, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.09343970566987991, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0187, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.1436399519443512, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0175, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.10790709406137466, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0184, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.14520230889320374, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0187, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.10826292634010315, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0188, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.1122521162033081, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0197, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.11445961147546768, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.015, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.1473265290260315, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0193, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.1572754830121994, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0186, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.1085842102766037, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0169, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.11982112377882004, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0163, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.12014015763998032, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.0217, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.11440391838550568, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0175, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.1276320368051529, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0203, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12434501200914383, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0186, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.11981593072414398, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0138, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.08389589190483093, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0162, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.11693917959928513, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.0199, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.1033058762550354, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0182, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.17878879606723785, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0181, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.12847506999969482, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0186, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.10514750331640244, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0166, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.1037413701415062, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0182, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.10402073711156845, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0165, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.10330315679311752, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0164, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.14082680642604828, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0205, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.1187078058719635, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0178, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.10590347647666931, + "learning_rate": 2.047995045386578e-05, + "loss": 0.0192, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.11178472638130188, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.019, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.11713241040706635, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0169, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.1520795077085495, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0173, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.11906655877828598, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0163, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.10854434967041016, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0181, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.16785599291324615, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.0212, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10305774211883545, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0179, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.09346644580364227, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0162, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11048682034015656, + "learning_rate": 2.044530211206255e-05, + "loss": 0.0158, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.14503270387649536, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0185, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.12631897628307343, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0182, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.14716260135173798, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0205, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.1364247053861618, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0187, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.09775770455598831, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0179, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.17999885976314545, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0194, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.1381615400314331, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0192, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.10749474912881851, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0189, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.15220865607261658, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.12083570659160614, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0175, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.11410468816757202, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0174, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.11285928636789322, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.0191, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.09808669239282608, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0173, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.07822877913713455, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0179, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.1032382920384407, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0177, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12453432381153107, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0183, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.1397278904914856, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0175, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.1701551228761673, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0176, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.13178907334804535, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0158, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.16403238475322723, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0182, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.13050618767738342, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.019, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09723208099603653, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.016, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.11937404423952103, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0194, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.10595990717411041, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0179, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.10598801076412201, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0172, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.12562280893325806, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0192, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.1062425896525383, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0167, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.19852802157402039, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0186, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10752324014902115, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0171, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.11313100904226303, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0191, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.09881386160850525, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0189, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.11926184594631195, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0176, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.11304130405187607, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0171, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.159687340259552, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0209, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.12324368953704834, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0184, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.10718217492103577, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0178, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12370229512453079, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0181, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.11843078583478928, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.017, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.11866020411252975, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0173, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.10999849438667297, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0217, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.10843008756637573, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.0163, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.13494807481765747, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0202, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.13634233176708221, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.0203, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.09910089522600174, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0154, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.1105995923280716, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0172, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.11397486180067062, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0171, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10141155868768692, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0187, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.09431210905313492, + "learning_rate": 2.025995044916193e-05, + "loss": 0.0163, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.09627058357000351, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0188, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.10498029738664627, + "learning_rate": 2.025220726516947e-05, + "loss": 0.0197, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.12692295014858246, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0163, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.11992908269166946, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0188, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.1448463350534439, + "learning_rate": 2.024058949838742e-05, + "loss": 0.0182, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.10764076560735703, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0191, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.1176103726029396, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0202, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.10393970459699631, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0174, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.12357908487319946, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0188, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11184806376695633, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0171, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09287009388208389, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0164, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.13145606219768524, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0175, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.108634814620018, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0192, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.1643940508365631, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.02, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.1642858386039734, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0195, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.11515326052904129, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0173, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.11438507586717606, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0187, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.1556805521249771, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0193, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.1081680878996849, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0176, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.11440131068229675, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0198, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.13953827321529388, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0163, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.12443852424621582, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0173, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.12884436547756195, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.0184, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.09170050173997879, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0158, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.08685769885778427, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0143, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.12880417704582214, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0172, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.09957600384950638, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0189, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.11359020322561264, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0191, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.10620398819446564, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0183, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.11106064170598984, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0175, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.14327868819236755, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0156, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.09743347764015198, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0165, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.11041118204593658, + "learning_rate": 2.013198534613597e-05, + "loss": 0.019, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.13757234811782837, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.017, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.10167638957500458, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0172, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.1386769413948059, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0157, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.13493874669075012, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0184, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.1238163560628891, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0178, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12875531613826752, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0176, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11199075728654861, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0179, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12402849644422531, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0197, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.12012362480163574, + "learning_rate": 2.009701166604228e-05, + "loss": 0.0203, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.1319619119167328, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0176, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.11875088512897491, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.017, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.14083963632583618, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0192, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1157609224319458, + "learning_rate": 2.008145773123833e-05, + "loss": 0.017, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1447380781173706, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0192, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.08668482303619385, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.0182, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.11417029052972794, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0162, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.12431302666664124, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0176, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.09466125816106796, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0174, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.13191643357276917, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0162, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.16682866215705872, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0167, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.09262735396623611, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0174, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10246162116527557, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0157, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.14533615112304688, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0193, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10985267907381058, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0185, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.1364467293024063, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0174, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.12792983651161194, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.019, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.13055621087551117, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0175, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.1272381991147995, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0191, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.1523309350013733, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.0192, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.12438428401947021, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0183, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.13417088985443115, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0192, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.10955093055963516, + "learning_rate": 2.000749254391884e-05, + "loss": 0.019, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.1140032708644867, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.0155, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.13144028186798096, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.0177, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.10048601031303406, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0169, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.1272922307252884, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0167, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.10301118344068527, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0166, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.10341407358646393, + "learning_rate": 1.998410649877165e-05, + "loss": 0.018, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.08479570597410202, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0193, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.10605745017528534, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0169, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.09806226938962936, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0175, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.12049946188926697, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0187, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.13633446395397186, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0167, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.12669390439987183, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.02, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.1395879089832306, + "learning_rate": 1.99568055617237e-05, + "loss": 0.02, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.1145915761590004, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.019, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12087064981460571, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0194, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.11252355575561523, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0156, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.10857690125703812, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0197, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.17054887115955353, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0172, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.12123709172010422, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0184, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.08856352418661118, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0178, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.09841718524694443, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0198, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.1236206591129303, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0171, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11047739535570145, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0198, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.11695606261491776, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0162, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.12013910710811615, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0171, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.08011548966169357, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0176, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.13469302654266357, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0177, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.12169095873832703, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.0192, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.1165102869272232, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0165, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.16436143219470978, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0164, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.11628631502389908, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0174, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.13279633224010468, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0164, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11530247330665588, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0199, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13026569783687592, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0193, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.12264525890350342, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0178, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.09320709109306335, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0157, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.11837475001811981, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.016, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.12293318659067154, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.0182, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.09163888543844223, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0197, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.1171371266245842, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0162, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.1096082329750061, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0198, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09364283084869385, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0164, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.08231621980667114, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0169, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.1273244321346283, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.0167, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.15531282126903534, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0194, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.13366153836250305, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0186, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.1072721853852272, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0173, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.11176268011331558, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0182, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.10074266046285629, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0167, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.11638782918453217, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0194, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.1109209880232811, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0197, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.1511884480714798, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0202, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.09859053790569305, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0185, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.12002220004796982, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0177, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.11930137127637863, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0159, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.10564088076353073, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0176, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.10428094118833542, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0197, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.1335987150669098, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0188, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.12865577638149261, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0179, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.14246998727321625, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0181, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.1250566691160202, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0188, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.09627831727266312, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0153, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.09008506685495377, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0185, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.12598916888237, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0169, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.11335116624832153, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0165, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.10588988661766052, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0171, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.10731689631938934, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.0174, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.11184538155794144, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0154, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.10870286822319031, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0163, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.11692065745592117, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0178, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.10036200284957886, + "learning_rate": 1.972989772246684e-05, + "loss": 0.016, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.12077154219150543, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0178, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.1272457093000412, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0188, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.08074570447206497, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0159, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.11381031572818756, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.0155, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.09836770594120026, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0161, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.09236133843660355, + "learning_rate": 1.970635454034925e-05, + "loss": 0.0183, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.1167517825961113, + "learning_rate": 1.970242942396647e-05, + "loss": 0.018, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.10953414440155029, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.015, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.12463351339101791, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0192, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.1285952925682068, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0172, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.10794857144355774, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0167, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.11815376579761505, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0176, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.10251157730817795, + "learning_rate": 1.967887123661664e-05, + "loss": 0.018, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.16941046714782715, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0182, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.14373992383480072, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.0177, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.11837214231491089, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0169, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.10599380731582642, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0167, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.11006144434213638, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0207, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.11699084937572479, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0164, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.10131341218948364, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0177, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.09975451976060867, + "learning_rate": 1.964744043284577e-05, + "loss": 0.017, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.08056963980197906, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0171, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.16518834233283997, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0186, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.10148414224386215, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0181, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.14867499470710754, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0183, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.16952896118164062, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0199, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13692231476306915, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0198, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.15133927762508392, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0184, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11434897035360336, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0189, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.0917418971657753, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0167, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12218283116817474, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0168, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.12193279713392258, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0173, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.09389539062976837, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0182, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.10455376654863358, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0169, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.11151516437530518, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0188, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.14512453973293304, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0194, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.10494859516620636, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0169, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.15839508175849915, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0177, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.12615326046943665, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0172, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.1135321632027626, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0166, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.12872013449668884, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0173, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.18429434299468994, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0165, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.13249747455120087, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0183, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.10603360831737518, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0145, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.10620120912790298, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0155, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.10986930131912231, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0163, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09298109263181686, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0179, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09945658594369888, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0178, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.08814677596092224, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0186, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.09042713046073914, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0173, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.11592321842908859, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0191, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.1026391014456749, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.018, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.09040834754705429, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0164, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.11391181498765945, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0177, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11113253980875015, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.0166, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.10178712755441666, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0167, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.13696520030498505, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0171, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.10290519148111343, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0164, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13472187519073486, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0173, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.08172356337308884, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0187, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.12297531962394714, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0166, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.13925355672836304, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0193, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.15394000709056854, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0169, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.09050899744033813, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.019, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.11854210495948792, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0181, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.10207716375589371, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0154, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.10006722062826157, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0161, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.10565274208784103, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0183, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.11656428128480911, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0166, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.12020296603441238, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0181, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.0964578241109848, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0174, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.0888240784406662, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0158, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.09271078556776047, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0187, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.10407985746860504, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0152, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1329568326473236, + "learning_rate": 1.94346993525268e-05, + "loss": 0.0166, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.13284213840961456, + "learning_rate": 1.943075029574398e-05, + "loss": 0.018, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.14158464968204498, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0194, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.11495806276798248, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0157, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.12886112928390503, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0174, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08954162895679474, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0158, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.09367296099662781, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0171, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.11188989877700806, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0162, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.10758768022060394, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0154, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.11783428490161896, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0182, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.18594273924827576, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0175, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.08010479807853699, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0167, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13949649035930634, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0175, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.10709669440984726, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0179, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.1675017923116684, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0176, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12764538824558258, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0165, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11073391884565353, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0169, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.10426445305347443, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0171, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.09000374376773834, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0161, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.10287991911172867, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0171, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.12322332710027695, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0178, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.11188322305679321, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0166, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.14056499302387238, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0196, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.08969348669052124, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0169, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.10100627690553665, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0159, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.1181580200791359, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0185, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.11281560361385345, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0171, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.11793895065784454, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0187, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.14534896612167358, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0177, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.10177207738161087, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0182, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.1335727721452713, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.017, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.14608687162399292, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0197, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.09763725847005844, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0184, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.09486788511276245, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0177, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.12235541641712189, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.018, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.13838450610637665, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0174, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.09837213158607483, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.0174, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.11713139712810516, + "learning_rate": 1.928836248385875e-05, + "loss": 0.016, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.13015268743038177, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0187, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.13498486578464508, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0181, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.12189928442239761, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0156, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.14399787783622742, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0185, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.10662715882062912, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0164, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.1145375519990921, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0185, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.11677944660186768, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0153, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.10242128372192383, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0168, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.1290024369955063, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0189, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.11820637434720993, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0167, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.12113048881292343, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0187, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.12122460454702377, + "learning_rate": 1.92408053936279e-05, + "loss": 0.0175, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.09413880109786987, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0149, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.11389796435832977, + "learning_rate": 1.923287468901755e-05, + "loss": 0.0185, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13923633098602295, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.0169, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.15947312116622925, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0188, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.14314432442188263, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0182, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.1426798403263092, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0179, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.12179753184318542, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.02, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11758018285036087, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0197, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11349253356456757, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0171, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.13724955916404724, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0181, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.1260363608598709, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0172, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.09541817009449005, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0181, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.0922083631157875, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0177, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.08051180839538574, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0167, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.11803245544433594, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0177, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10018756985664368, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.10512874275445938, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0164, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.17478078603744507, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0169, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.08864858746528625, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0165, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11313372850418091, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0172, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.09432164579629898, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0163, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.09435626119375229, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0188, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.10041314363479614, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0157, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.07982632517814636, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0176, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.09699293971061707, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0177, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.11059583723545074, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.017, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.11384644359350204, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0163, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.11473467200994492, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0165, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.11018316447734833, + "learning_rate": 1.912568530871305e-05, + "loss": 0.0179, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.10781396180391312, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0199, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.11702359467744827, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0165, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.12909658253192902, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0162, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.09839031100273132, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0152, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.1357898861169815, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0198, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.11221814155578613, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.0178, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.11431093513965607, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0176, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.11436092853546143, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0187, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.09838918596506119, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0148, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.11489785462617874, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0168, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.10876795649528503, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0169, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.1032189428806305, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0152, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.10013218969106674, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.11554815620183945, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0182, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.2529490292072296, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0198, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.09215852618217468, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0168, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.12875677645206451, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0185, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.1232401430606842, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0179, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.10334797203540802, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0171, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10524798929691315, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0162, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.11014945805072784, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0176, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.1074177473783493, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0162, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.14043818414211273, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0191, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.10011711716651917, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0178, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0172, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.12526027858257294, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0137, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.15774008631706238, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0153, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.1049765795469284, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0155, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09307108074426651, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0186, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1369069367647171, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0158, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.12320058792829514, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0181, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.10885081440210342, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0168, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.08989177644252777, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0153, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.09219451993703842, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0172, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.11661440879106522, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0192, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.0971342995762825, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0178, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.10453730076551437, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0165, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.17652514576911926, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0196, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10733854025602341, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0154, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.0989762544631958, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0181, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.12107723951339722, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.1385297179222107, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0191, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.12209209054708481, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0217, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.09332361072301865, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0158, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.12448447197675705, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0164, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.12061425298452377, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0169, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.12132656574249268, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0183, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.14007116854190826, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0167, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.11448334902524948, + "learning_rate": 1.893057716200171e-05, + "loss": 0.015, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.12702956795692444, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0152, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.09236849099397659, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0177, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.09818986803293228, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.0164, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12131274491548538, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0171, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10609138011932373, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.0147, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.1431477814912796, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0173, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.0844314768910408, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0162, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.09800038486719131, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0157, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.11529140919446945, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0172, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.086520716547966, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0154, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.12068255245685577, + "learning_rate": 1.888667760339568e-05, + "loss": 0.015, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.12278824299573898, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0174, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.10242115706205368, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0173, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.11696609109640121, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0187, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.11133866757154465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0151, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.1319858729839325, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0165, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.08355733752250671, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0166, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.10663259029388428, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0177, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.09787473827600479, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0154, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.12111647427082062, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0171, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.10036861151456833, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.15351398289203644, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.019, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.11561891436576843, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0174, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.09197179228067398, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0174, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.0961032509803772, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.019, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.09264974296092987, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0158, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.11056753247976303, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0159, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.10464371740818024, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0165, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.13888190686702728, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0185, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.09801805019378662, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0189, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.15648017823696136, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0167, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1129915863275528, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0166, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.10456246137619019, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0171, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.10873697698116302, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.0168, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.17470459640026093, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0169, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.15504027903079987, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.10821744054555893, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0152, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.14031334221363068, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0181, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.128667414188385, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0183, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.09095419198274612, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0151, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.09303424507379532, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0167, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.10706666111946106, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0174, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.13011431694030762, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0177, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.13821059465408325, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0154, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.10680826753377914, + "learning_rate": 1.875076403503638e-05, + "loss": 0.017, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.12847846746444702, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0146, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.1472046673297882, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0184, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10622651875019073, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0153, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.129075825214386, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0189, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.12048091739416122, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.017, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.10954522341489792, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0175, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.10811172425746918, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0173, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.13378627598285675, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0193, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.13185591995716095, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0195, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.1013568863272667, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0182, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.07558644562959671, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0173, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.10668148100376129, + "learning_rate": 1.870271530813327e-05, + "loss": 0.018, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.134077250957489, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0157, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.14325739443302155, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0221, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.1389995515346527, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0178, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.13043907284736633, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0169, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.12082460522651672, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0174, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11521385610103607, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0165, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.11550573259592056, + "learning_rate": 1.867466815081593e-05, + "loss": 0.017, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.11315014213323593, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0173, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.1067388728260994, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0161, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.0941673144698143, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0158, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.09685055166482925, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0161, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.11744765937328339, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0164, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.10552993416786194, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0154, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.12654545903205872, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0184, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.11333387345075607, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0186, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.09856864809989929, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.0187, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.0933985561132431, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0171, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.096091628074646, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0194, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.08463368564844131, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0165, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.10648614913225174, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.017, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.10691184550523758, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0165, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1005169153213501, + "learning_rate": 1.861452118736741e-05, + "loss": 0.0174, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.1037023589015007, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0172, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.1299487054347992, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.015, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14639562368392944, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0167, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.12270855903625488, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0173, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.09706515073776245, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0169, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.11888926476240158, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.0152, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.11289773136377335, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0175, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13422279059886932, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0168, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.10992032289505005, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0176, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.10719195753335953, + "learning_rate": 1.85743888359202e-05, + "loss": 0.017, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.10348568856716156, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0166, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.10436776280403137, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0172, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.09779159724712372, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.018, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.10799746960401535, + "learning_rate": 1.855832827763672e-05, + "loss": 0.016, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.11972185224294662, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0195, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.10282087326049805, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0169, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.11840159446001053, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0183, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.13861460983753204, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0205, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10774466395378113, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0163, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.14446309208869934, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0168, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.1313662976026535, + "learning_rate": 1.853021190846946e-05, + "loss": 0.0161, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.0857798308134079, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0159, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.11933837085962296, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0178, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11762946099042892, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0154, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.11128528416156769, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0198, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.07360108196735382, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0148, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.10680225491523743, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0174, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.11387799680233002, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0188, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.2774786651134491, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0165, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.12314122170209885, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.017, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.14403507113456726, + "learning_rate": 1.849002290456207e-05, + "loss": 0.0175, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.10811382532119751, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0173, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.1399172991514206, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0154, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.10577643662691116, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0177, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.10794928669929504, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0155, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.0844842791557312, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0149, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.09744124114513397, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0166, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.14749610424041748, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0179, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.09801949560642242, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0177, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.07795082777738571, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0174, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.11197780072689056, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.0172, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.11819472163915634, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.08461402356624603, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0171, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.11102861911058426, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0156, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.11057187616825104, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0189, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.11031147837638855, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0155, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.08435497432947159, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0187, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.12450679391622543, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.017, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.09348208457231522, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0171, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.10584745556116104, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0161, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.14194349944591522, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.0177, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.1025288999080658, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0153, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.07800278812646866, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0157, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.1253945231437683, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0176, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.11724173277616501, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.016, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.13043718039989471, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0179, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.08728159964084625, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0177, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.1419975906610489, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0185, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.12774096429347992, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0187, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.10763564705848694, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.0163, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.12671130895614624, + "learning_rate": 1.836929808825424e-05, + "loss": 0.0177, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.10261370241641998, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0145, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12106029689311981, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.0157, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.12428469955921173, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0194, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.08797722309827805, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0175, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.14663904905319214, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.018, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.12288253754377365, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0176, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.11406450718641281, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0164, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10584931075572968, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.017, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1212797686457634, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0176, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.10806597024202347, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0159, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.10098059475421906, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0186, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.14030125737190247, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0204, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.09907040745019913, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0158, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1299905776977539, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0198, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.134843647480011, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0157, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.1048227846622467, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0154, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.12262892723083496, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0193, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.12806116044521332, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0167, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.12875764071941376, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.0149, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.09743807464838028, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0149, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10791302472352982, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0155, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10845673829317093, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0187, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.1366390436887741, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0161, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.101824089884758, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0174, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.12938271462917328, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0159, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.13503864407539368, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0196, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10241002589464188, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0153, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.10051950812339783, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.0159, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11316224932670593, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0166, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.10802372545003891, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0167, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.1365254521369934, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0176, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.11756277829408646, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.016, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13363559544086456, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0167, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.13138234615325928, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0194, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.1519714891910553, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0147, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.12152288109064102, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.017, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.0976492166519165, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0174, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.09356466680765152, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0158, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.09382101893424988, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0175, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.12188217788934708, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0163, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.09245157241821289, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0183, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.1661074161529541, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0202, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.154639333486557, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0143, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.117487333714962, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.017, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.09468945115804672, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0171, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.09632660448551178, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0147, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.12390028685331345, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0177, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.11924377828836441, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0154, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.13191300630569458, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0154, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.07403766363859177, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0171, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.10911701619625092, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0159, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.11888141930103302, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0189, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.09766717255115509, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0182, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.08285287767648697, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0145, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.13635197281837463, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0166, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.11627981066703796, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.10899969190359116, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0174, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.10002870857715607, + "learning_rate": 1.813525107780979e-05, + "loss": 0.016, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09963744133710861, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0163, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.2032729536294937, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0188, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.10394410789012909, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0178, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.1004093736410141, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0181, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.08776404708623886, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.0188, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.10926580429077148, + "learning_rate": 1.811099258182467e-05, + "loss": 0.018, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.09496597945690155, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0153, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.12430521845817566, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0152, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.11681065708398819, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0156, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.10450825840234756, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0162, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10932883620262146, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0163, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.1358717978000641, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0152, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.11744217574596405, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0188, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.14966122806072235, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0179, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.113619364798069, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.08801224827766418, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0168, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.09556800872087479, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0171, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.11116721481084824, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0171, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.10650353878736496, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0158, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.09222083538770676, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0154, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.10958139598369598, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.0189, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11020398885011673, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0166, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.08026755601167679, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0159, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.09629099816083908, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0162, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11310935020446777, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0192, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.1281706690788269, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0191, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12402708828449249, + "learning_rate": 1.802602130181036e-05, + "loss": 0.0159, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.09760920703411102, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0151, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.12342818826436996, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.018, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09423290938138962, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0163, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.10432913154363632, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0172, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.09628765285015106, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0169, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.11590366810560226, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0167, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10148236900568008, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0155, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.10599179565906525, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0155, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.11854606866836548, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.10626410692930222, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0185, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.12172172963619232, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0179, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.10842256247997284, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0173, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.11806859076023102, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0163, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11489225924015045, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0178, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.11859364807605743, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0164, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.18860413134098053, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.0168, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.13375963270664215, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0168, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.13163933157920837, + "learning_rate": 1.795310800633073e-05, + "loss": 0.0151, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.11135797202587128, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0175, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.12392686307430267, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0161, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.0995136946439743, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0165, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.10128366202116013, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0161, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.13080070912837982, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0169, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.10262267291545868, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0176, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.10983351618051529, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0163, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.09834928810596466, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0158, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.11170317232608795, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0166, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.14308896660804749, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.09110505878925323, + "learning_rate": 1.790851399590954e-05, + "loss": 0.017, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.12270113825798035, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0165, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.1422351896762848, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0177, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1294422596693039, + "learning_rate": 1.789634733729274e-05, + "loss": 0.017, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.1157895103096962, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0192, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.12473416328430176, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0156, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.1765025109052658, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0161, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11862069368362427, + "learning_rate": 1.78801220478228e-05, + "loss": 0.017, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.10549142211675644, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0165, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.09818614274263382, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0171, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.13607585430145264, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0168, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.09340739995241165, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0158, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.0805685818195343, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.0207, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.11172603815793991, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0164, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.08757263422012329, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0179, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.09590335935354233, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0199, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.1186867505311966, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.0146, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.09467568248510361, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0175, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11578786373138428, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0179, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09604159742593765, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0173, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.11337848007678986, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.0156, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.11909373849630356, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0188, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.08804072439670563, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0175, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.13454975187778473, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.0184, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.12201394885778427, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0148, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.11946159601211548, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0153, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12322071939706802, + "learning_rate": 1.78030044678245e-05, + "loss": 0.0181, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.10444212704896927, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0159, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.10723353922367096, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0158, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.10750993341207504, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0167, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.1207774356007576, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0207, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.12730205059051514, + "learning_rate": 1.778269750931628e-05, + "loss": 0.017, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.08926276862621307, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0151, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.10322589427232742, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0174, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10034044831991196, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.0184, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.12131599336862564, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0183, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.10715801268815994, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0156, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.14093570411205292, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0169, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.124595507979393, + "learning_rate": 1.775425890028875e-05, + "loss": 0.017, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.13153409957885742, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0189, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.1348494291305542, + "learning_rate": 1.77461316962853e-05, + "loss": 0.019, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.10926292091608047, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0161, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.11094044893980026, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0155, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.1341349333524704, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0184, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.15732111036777496, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0203, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.07597270607948303, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.0184, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.11097310483455658, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.0176, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.12087065726518631, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0164, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12998445332050323, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0191, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.13928557932376862, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0157, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.10487896203994751, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0172, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.13437892496585846, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0182, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.20842674374580383, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0189, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.09813268482685089, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12833304703235626, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0168, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.1019122451543808, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0159, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.07276969403028488, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.016, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09712188690900803, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0169, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.10403666645288467, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0156, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.11877460777759552, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0152, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.10731421411037445, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0188, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.12387207895517349, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0173, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.10561594367027283, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0179, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.2772648334503174, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0168, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.097810760140419, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0142, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.1168239563703537, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0171, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.14195922017097473, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0162, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11385945230722427, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0187, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.1124262586236, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.13074515759944916, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0151, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.12591321766376495, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0177, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.09103676676750183, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0169, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.12457307428121567, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0162, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.11226223409175873, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0156, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.14778819680213928, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0187, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.10433483868837357, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0187, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.08798553794622421, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0163, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.09729046374559402, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0149, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10188018530607224, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0154, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.12559989094734192, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.018, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12567821145057678, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0168, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.12211210280656815, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0166, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.11825426667928696, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0168, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.11676980555057526, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0175, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.11447525024414062, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0162, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.08958559483289719, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0161, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.11528483778238297, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0162, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.08296159654855728, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.014, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.12306056916713715, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0158, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.13315193355083466, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.0164, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.09533339738845825, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0166, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.20049314200878143, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.017, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.14355438947677612, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0187, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.11443490535020828, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0146, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.10336155444383621, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0164, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11926902830600739, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0173, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.09699229151010513, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0176, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.11406200379133224, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0164, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.10817506164312363, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0165, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.11398246139287949, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.0163, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.09427376091480255, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0166, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.10372194647789001, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0148, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.09003781527280807, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0172, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.09807098656892776, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.11895686388015747, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0163, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10839524865150452, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0173, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.20792271196842194, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0183, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.07746267318725586, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0147, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.11470094323158264, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0177, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.1377800703048706, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.1247936338186264, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.016, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.1405080407857895, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0164, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.1318410336971283, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0175, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.14254692196846008, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0159, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.12803269922733307, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0146, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11180103570222855, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0169, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.16513720154762268, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0162, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.13401633501052856, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.0182, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.13510771095752716, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0175, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.12391090393066406, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0174, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1019999086856842, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0157, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.11906205862760544, + "learning_rate": 1.7416306237948e-05, + "loss": 0.0179, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11846558749675751, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.0154, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.12292807549238205, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0163, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10195571929216385, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0186, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.1032365933060646, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0163, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1241045668721199, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0173, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.10956405848264694, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.0163, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.12094716727733612, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.018, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.1353612244129181, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0161, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.0935104638338089, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0164, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.10712502896785736, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.0177, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.12715180218219757, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0148, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.09601753205060959, + "learning_rate": 1.736733684019633e-05, + "loss": 0.0167, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.11189524084329605, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0174, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.11025495082139969, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.016, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.1203494742512703, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0192, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.12150877714157104, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0143, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.11957377195358276, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0146, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.16482087969779968, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0181, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.10647732019424438, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0178, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.13345423340797424, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0155, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.13268540799617767, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0178, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.12209953367710114, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0161, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.11951886862516403, + "learning_rate": 1.73224254699117e-05, + "loss": 0.0213, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.12044327706098557, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0195, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.10611505806446075, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0169, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.12347753345966339, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0145, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13930706679821014, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0185, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.12702012062072754, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0167, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.1193593367934227, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0169, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.16778719425201416, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0179, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11699119210243225, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0159, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11571445316076279, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.08481728285551071, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0159, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.10007977485656738, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0155, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.12851299345493317, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0167, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.13929468393325806, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0154, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.11237664520740509, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0143, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.11654951423406601, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0179, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.11065302044153214, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0168, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.10024461150169373, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0173, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.1334248036146164, + "learning_rate": 1.724888829234145e-05, + "loss": 0.016, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.1115458682179451, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0159, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.15415695309638977, + "learning_rate": 1.72407140459686e-05, + "loss": 0.0166, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10965333878993988, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0159, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.10152038931846619, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0152, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.10030021518468857, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0161, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.1526176780462265, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0134, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13154515624046326, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.017, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.12260743975639343, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0152, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.07395336031913757, + "learning_rate": 1.721209883968656e-05, + "loss": 0.0155, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.1107548177242279, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0178, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.08634423464536667, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0148, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.10544566810131073, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0151, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.13896141946315765, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0171, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10351476818323135, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0164, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.11912661790847778, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0163, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.12101101130247116, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0185, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.09519081562757492, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0137, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.13252407312393188, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0153, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.08964477479457855, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0148, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.11772924661636353, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0153, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1261531263589859, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.0155, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.12295454740524292, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0162, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.14834828674793243, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0162, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.14458344876766205, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0146, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.10006255656480789, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0183, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11527495086193085, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0174, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.11176037788391113, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0146, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.11118462681770325, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0139, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.1216764822602272, + "learning_rate": 1.713029611235338e-05, + "loss": 0.018, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10572952032089233, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0155, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.09299932420253754, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0193, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.11654805392026901, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0146, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.125601127743721, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0148, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.12143341451883316, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0144, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.10944563895463943, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0188, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.10531438142061234, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.017, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.08527344465255737, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0134, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.12489438056945801, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0184, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.12359192967414856, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0162, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.09504237025976181, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0147, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.1144440770149231, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0155, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.11942129582166672, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.0161, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.15496809780597687, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0167, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.11641886830329895, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0155, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.13092041015625, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0164, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.1662251204252243, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0169, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.1151951253414154, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0146, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.13008327782154083, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.017, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.08816810697317123, + "learning_rate": 1.70484286724112e-05, + "loss": 0.0147, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.12602686882019043, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0167, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10360333323478699, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0161, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.10288142412900925, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0174, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.1306314915418625, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0178, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.11859587579965591, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0152, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.12666994333267212, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.0155, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.10859639197587967, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0174, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.1423543244600296, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0145, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.12444380670785904, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0165, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.07491649687290192, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0174, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.12173314392566681, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0156, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.11790800094604492, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.1375807374715805, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0182, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.11894135922193527, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0149, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.09377862513065338, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0157, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1448003053665161, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0189, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.14296171069145203, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.019, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.10862060636281967, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0188, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.1081303209066391, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.0166, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08262617141008377, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0163, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12013579159975052, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0161, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.06014449894428253, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0166, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09688957780599594, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0164, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.11730696260929108, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.0142, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.11243025213479996, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0134, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11974845081567764, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0156, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.1408282220363617, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0193, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.1495579332113266, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.017, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.1408172994852066, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.19624975323677063, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0165, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09940195083618164, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0168, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.11277889460325241, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0171, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.13027498126029968, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0165, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.11570801585912704, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.0159, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.12621626257896423, + "learning_rate": 1.690501241902545e-05, + "loss": 0.016, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.13131381571292877, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0146, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.11577112972736359, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0168, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.1240323930978775, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0162, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.17400938272476196, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.0155, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.09539356827735901, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0165, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10108646005392075, + "learning_rate": 1.688040861067922e-05, + "loss": 0.016, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.1295158863067627, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0151, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.13638637959957123, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0188, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.12255847454071045, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0148, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.13242852687835693, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0166, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11082842200994492, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.015, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.14538930356502533, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0172, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1609591394662857, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0166, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.10611626505851746, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.0142, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.13291358947753906, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0179, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.1112443208694458, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0179, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.10251478105783463, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0174, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.10443990677595139, + "learning_rate": 1.683118563837068e-05, + "loss": 0.016, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.08651736378669739, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0176, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.09996092319488525, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.017, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.13269121944904327, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.0154, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.121453657746315, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0167, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.10587169975042343, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0161, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.12990498542785645, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0158, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.12159401923418045, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0158, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.12046501785516739, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0165, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.10626068711280823, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0144, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.14602169394493103, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0161, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09537681937217712, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0188, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.1014745756983757, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0171, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.11358807235956192, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0165, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.12290225178003311, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0153, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.11178604513406754, + "learning_rate": 1.676962882515376e-05, + "loss": 0.0175, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.1464385688304901, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0155, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.12778854370117188, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0144, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.09556689858436584, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0171, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.10866247862577438, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.0151, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.14120467007160187, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0163, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.12688933312892914, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0174, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.10251957923173904, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0171, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.11465741693973541, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.016, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.08181405812501907, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0201, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.09379345923662186, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0172, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.12122894823551178, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0183, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.0956249088048935, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0142, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.11953603476285934, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0169, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.12494538724422455, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0167, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.20808497071266174, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0138, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.10776039958000183, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0152, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.09606330841779709, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0154, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.10988017171621323, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0182, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.10328707098960876, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0146, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.166867196559906, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0179, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.11701484769582748, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0159, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.10510280728340149, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0179, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.12857040762901306, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0172, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1627577394247055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0176, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.10962127149105072, + "learning_rate": 1.666696746039845e-05, + "loss": 0.0147, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.12961529195308685, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0164, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.08055070787668228, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0148, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.10041354596614838, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0161, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.12196766585111618, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0138, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.0879533439874649, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0156, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.1102190688252449, + "learning_rate": 1.664231677578745e-05, + "loss": 0.016, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.12463485449552536, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.016, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.10829959809780121, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0181, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.08602527529001236, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0176, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.10241527855396271, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.0172, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11007986962795258, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0156, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13110534846782684, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.018, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1417531818151474, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0161, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.13233700394630432, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0147, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.09148544073104858, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0145, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.08781567215919495, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.0149, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.10577751696109772, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0161, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.11689155548810959, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0171, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.0777219608426094, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0152, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.07983718812465668, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0158, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1595761626958847, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0168, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.1037725880742073, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0161, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.10239970684051514, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0149, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.15451422333717346, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0199, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.11882807314395905, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0165, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.09850180149078369, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0147, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.10899021476507187, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.09254994243383408, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0148, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.13559140264987946, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0162, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.11541012674570084, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0154, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.07937203347682953, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0153, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.10062690824270248, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0165, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.13341091573238373, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0178, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.0857294499874115, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0132, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.10551585257053375, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0142, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.12363193184137344, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0165, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.11550596356391907, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0165, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.09379533678293228, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0154, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.10079154372215271, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0163, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.13103018701076508, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0178, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.1172979474067688, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0152, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.11038312315940857, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0152, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.0925338938832283, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0148, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.2519476115703583, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.017, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.09703299403190613, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0193, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.10749004036188126, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0175, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.08736859261989594, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0147, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.13125599920749664, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0161, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10677823424339294, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0156, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.09135954827070236, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0151, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.10548236966133118, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0179, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.1150064468383789, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0188, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.127374529838562, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.016, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.09152884781360626, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0177, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.1021062359213829, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0145, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.1054648831486702, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0172, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.12128820270299911, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0165, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.11375363171100616, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0151, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.11004973948001862, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0182, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.11101444810628891, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0164, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.10393026471138, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0172, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.1139075756072998, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0159, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1569976508617401, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0178, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09649650752544403, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0165, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.16198694705963135, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0157, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.10348767787218094, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0173, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.12259576469659805, + "learning_rate": 1.639145830609088e-05, + "loss": 0.0171, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.10480271279811859, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0159, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.11254257708787918, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0163, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.11534930765628815, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0149, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.10350872576236725, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0169, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.10812029242515564, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.015, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.11796360462903976, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.0144, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.1276526153087616, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0157, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.11665219813585281, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0151, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.10940699279308319, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0165, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.09596745669841766, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0149, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15527229011058807, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0164, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11651212722063065, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0145, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.12004360556602478, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0153, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.13298730552196503, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0176, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.1496417224407196, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0153, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.11104541271924973, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0158, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.12215209007263184, + "learning_rate": 1.632147383062092e-05, + "loss": 0.015, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.09335075318813324, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0148, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.11481857299804688, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0153, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14369069039821625, + "learning_rate": 1.630912057457699e-05, + "loss": 0.019, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.11009083688259125, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0162, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.11112431436777115, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.0137, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.10135089606046677, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0167, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09966760873794556, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0168, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12062226980924606, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0145, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.1109282448887825, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0162, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.10862662643194199, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0152, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.09910363703966141, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0156, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.1374131590127945, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0167, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.12114953994750977, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.0156, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.14391179382801056, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0156, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.12364984303712845, + "learning_rate": 1.625969868712804e-05, + "loss": 0.0169, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.08771048486232758, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0142, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.10959397256374359, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0176, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.138074591755867, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0149, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.1116161122918129, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0154, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.1398090422153473, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0164, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.12145344167947769, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0173, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.11162104457616806, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0153, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.1211397722363472, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.016, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09488996863365173, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0146, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.11799386143684387, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0137, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.1406150758266449, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.0149, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.1307811290025711, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0192, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.13287024199962616, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.0153, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10396360605955124, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0133, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.11891327053308487, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0167, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.12922705709934235, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0178, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10563277453184128, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0132, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.17530488967895508, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0159, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.10669656842947006, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0166, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.09745346754789352, + "learning_rate": 1.617729852865125e-05, + "loss": 0.0152, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.12519428133964539, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0136, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.10747594386339188, + "learning_rate": 1.616905650452066e-05, + "loss": 0.0147, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.10352121293544769, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0161, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.1046888679265976, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0131, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.2108439952135086, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0148, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.1401670128107071, + "learning_rate": 1.61525713933778e-05, + "loss": 0.0171, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.10765372961759567, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0169, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.14557960629463196, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0151, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.11173100024461746, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0181, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.08689062297344208, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0156, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.10052058845758438, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0171, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.11383800208568573, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0167, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10390642285346985, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0157, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.10242070257663727, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0147, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.10699810087680817, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.0162, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.1349048614501953, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0167, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.10787097364664078, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.0144, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.06869964301586151, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0168, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.08449438959360123, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.0156, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10576694458723068, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0157, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.10042312741279602, + "learning_rate": 1.609073991709496e-05, + "loss": 0.018, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.10223086178302765, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0179, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.11301153153181076, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0151, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.11257045716047287, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0159, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.09129524976015091, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0154, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.10601308941841125, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0157, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.11378258466720581, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0151, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.1138806939125061, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0188, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.10367786139249802, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0157, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.13393838703632355, + "learning_rate": 1.605363201758077e-05, + "loss": 0.016, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.09532077610492706, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0145, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.13368268311023712, + "learning_rate": 1.60453849294712e-05, + "loss": 0.0178, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10675307363271713, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0148, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.10722523182630539, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0167, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.09825004637241364, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0148, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.14284753799438477, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.0194, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0167, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.13374702632427216, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0153, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.08964364975690842, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.0164, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.07354595512151718, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0138, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.08927512913942337, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0167, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.10888224840164185, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0193, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.12273052334785461, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0157, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.10807470232248306, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0152, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.1124584972858429, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.0148, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.10052672028541565, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0145, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.08666819334030151, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0172, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.11364969611167908, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0155, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.11263672262430191, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0164, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.1379120796918869, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0187, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.10857756435871124, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0133, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.13408084213733673, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0156, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.10847405344247818, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0141, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.10151072591543198, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0138, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11481117457151413, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0138, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.10202853381633759, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0153, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.09370319545269012, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.0159, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.09684473276138306, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0162, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.11064363270998001, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0162, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.15720140933990479, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0167, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.12132807075977325, + "learning_rate": 1.592576773310894e-05, + "loss": 0.0159, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.10888595134019852, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0133, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.0892426148056984, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0163, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12278196215629578, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0186, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10831111669540405, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0155, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.08722463995218277, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0156, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.11901626735925674, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0155, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.08866825699806213, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.0155, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.09303046017885208, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0152, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.12453419715166092, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0166, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.11046329140663147, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0155, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.10482461750507355, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0154, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11235640943050385, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0165, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.11749114841222763, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0159, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.1201302632689476, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0174, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.13044941425323486, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0143, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.10600301623344421, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0155, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.12556208670139313, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0159, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.12454571574926376, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0153, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.13055634498596191, + "learning_rate": 1.584736531020125e-05, + "loss": 0.0188, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.1379219889640808, + "learning_rate": 1.584323820456826e-05, + "loss": 0.016, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.13937926292419434, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0181, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.21694405376911163, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.013, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.11387006938457489, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0144, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.10102641582489014, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0154, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.10343687981367111, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0161, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.14018288254737854, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0175, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10792441666126251, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0168, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.10692388564348221, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0175, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.11248136311769485, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0176, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.13729289174079895, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0154, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.1389625370502472, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0152, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11402085423469543, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0154, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.14007720351219177, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0151, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.10529543459415436, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0151, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.09746415913105011, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0136, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1150866374373436, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0158, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.10207786411046982, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0139, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.13951019942760468, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0157, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11709082126617432, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.0158, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.12401671707630157, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0145, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.103446826338768, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0149, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.16494140028953552, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0192, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.10893384367227554, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0141, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.09813039004802704, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.0142, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.14741374552249908, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0184, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.12035968154668808, + "learning_rate": 1.573591189751e-05, + "loss": 0.0157, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.0970459058880806, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0176, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.10946980863809586, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0144, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.15628819167613983, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0157, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.12485401332378387, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0184, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.11361467838287354, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0163, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.19872187077999115, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0152, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.0912356749176979, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0167, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.09560416638851166, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0178, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.10187338292598724, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0127, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.11934805661439896, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0145, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.11719757318496704, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0148, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09480439126491547, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0167, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.09031171351671219, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0163, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.12236128002405167, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.014, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.09773887693881989, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.015, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.11999719589948654, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.0145, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.11341046541929245, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0142, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.10415472090244293, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0183, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.10177286714315414, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0152, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.10778342932462692, + "learning_rate": 1.565332720308093e-05, + "loss": 0.016, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.08542678505182266, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0157, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.10727313905954361, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0164, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.1413055658340454, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0163, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.10777363926172256, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0175, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.14286576211452484, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0155, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.10413800925016403, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0164, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.10655509680509567, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0163, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.10989648103713989, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0153, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.09240605682134628, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0144, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.0842762365937233, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0143, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.13544097542762756, + "learning_rate": 1.560789700123175e-05, + "loss": 0.017, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11756332218647003, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0159, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.11726696044206619, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0165, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.11693235486745834, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0146, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.1371903419494629, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0145, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.0971926674246788, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.017, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.08220933377742767, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0156, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.10194069892168045, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0134, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.08653450012207031, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.0131, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.08990923315286636, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0144, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.10370788723230362, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0133, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.06916087865829468, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.0139, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.1136801689863205, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0146, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12992094457149506, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.0144, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.09605134278535843, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0165, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.10273651778697968, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0165, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.11571857333183289, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0151, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.10102164000272751, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0146, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.11045269668102264, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.0146, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.12626338005065918, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0166, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.1357768476009369, + "learning_rate": 1.552528261458322e-05, + "loss": 0.018, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.10587792843580246, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0143, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.1509760171175003, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0187, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11034028977155685, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0174, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.07888679206371307, + "learning_rate": 1.550875774213821e-05, + "loss": 0.0159, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.11219298094511032, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.13006654381752014, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.0161, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12485881894826889, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0148, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.10320480912923813, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0171, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.12367358803749084, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.019, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1412898451089859, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0196, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.10843265801668167, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.0146, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09489759802818298, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0147, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.10236489772796631, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0158, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.12652312219142914, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0173, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10095653682947159, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0167, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.0802559182047844, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0145, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.1239897757768631, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0155, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.09682597219944, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.0154, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.10482888668775558, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.013, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.1272023320198059, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0164, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.09988776594400406, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0166, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.10352584719657898, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0142, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.36612969636917114, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0169, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.12097200751304626, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09437882900238037, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.0147, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14130210876464844, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0147, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.1284940391778946, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0148, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12240798771381378, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0135, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.08258320391178131, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0154, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.10075270384550095, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0162, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.10424498468637466, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0138, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.10447008907794952, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0156, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.10173123329877853, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0127, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.12066177278757095, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0162, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.09489274024963379, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0159, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.09174267202615738, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0155, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.13935479521751404, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0158, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.11126208305358887, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0142, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.13917426764965057, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.014, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.10231155157089233, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0141, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.11477960646152496, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0145, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.10886958241462708, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0155, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.09005019068717957, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0147, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.09915502369403839, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0138, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.1025547906756401, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0158, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.1030225157737732, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0158, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.0842776820063591, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0163, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.0634579285979271, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0151, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.09493366628885269, + "learning_rate": 1.532281505788039e-05, + "loss": 0.014, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12343619018793106, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0167, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.10681836307048798, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0149, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.1302853226661682, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0171, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.10506479442119598, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0136, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.12051218003034592, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0168, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11805758625268936, + "learning_rate": 1.52980183248759e-05, + "loss": 0.0158, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.10309842228889465, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.015, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.108099564909935, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0159, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11023534834384918, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0136, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.11571311205625534, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0163, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.1068359762430191, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.017, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.15622149407863617, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0156, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.11074791103601456, + "learning_rate": 1.526908777801797e-05, + "loss": 0.0163, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.11449336260557175, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0156, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.1219671219587326, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.0167, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.08449103683233261, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0162, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.13202930986881256, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0167, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.11665860563516617, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0173, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.13584332168102264, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0148, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.09913904964923859, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0155, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09704537689685822, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0162, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.13063789904117584, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0153, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.1202530562877655, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0182, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.08398830890655518, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.015, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.09059721976518631, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0129, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.10968997329473495, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0141, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.09898033738136292, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.1568402498960495, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0161, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.09885324537754059, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0168, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09452448785305023, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0142, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.16158591210842133, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0149, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.13292880356311798, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.016, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.17052333056926727, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0165, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.11351195722818375, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0163, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.09296253323554993, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0128, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.09396881610155106, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0157, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.13142332434654236, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0161, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.1246478259563446, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0135, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.13347554206848145, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0173, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.10838479548692703, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.11035843938589096, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0158, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.10406306385993958, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0149, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.16070416569709778, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0149, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09154598414897919, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.014, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.10438812524080276, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0164, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.09018818289041519, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0149, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1320919543504715, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0151, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.0870271846652031, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0155, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.12103381007909775, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0191, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.09007984399795532, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.0141, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.10315215587615967, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.19641031324863434, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0169, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14303788542747498, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0207, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.10933230072259903, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0136, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.11214675009250641, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0175, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.09008116275072098, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0145, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.09065063297748566, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.016, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.1116153672337532, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.016, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.09231071174144745, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0148, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12521710991859436, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0159, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.12240038067102432, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0143, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.09896460920572281, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0154, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.10471562296152115, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0152, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.11139113456010818, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0132, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.09953293949365616, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.015, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.12128512561321259, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0155, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.14064094424247742, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0132, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.09164810925722122, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0144, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.11127574741840363, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0145, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.11218927800655365, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0168, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.10342293232679367, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0137, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.10499373823404312, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0151, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.08029376715421677, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0157, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09445271641016006, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0161, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.09650483727455139, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0155, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.08300688117742538, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0176, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.11890654265880585, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0145, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10451909154653549, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0195, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.08795508742332458, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0141, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.1659373790025711, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0197, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.1001206561923027, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0161, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10190606862306595, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0161, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.11775147914886475, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0139, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.11382468789815903, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0149, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10219351947307587, + "learning_rate": 1.497147766283292e-05, + "loss": 0.0143, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.0930316224694252, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0148, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.10227128118276596, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0153, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.10634683817625046, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0163, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.1443759799003601, + "learning_rate": 1.495494301522528e-05, + "loss": 0.016, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.10867374390363693, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0147, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.11999388039112091, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0157, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10372292995452881, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0161, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.14905284345149994, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.0172, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10412498563528061, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0145, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.09797366708517075, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0133, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.1329793930053711, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0179, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.09716201573610306, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0146, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.09868425875902176, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.0153, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.0929884985089302, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0159, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.1417207568883896, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0151, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.13194335997104645, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0177, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.11850415170192719, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0173, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08122166991233826, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0149, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.11195199191570282, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.12405557930469513, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0135, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.12344473600387573, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0153, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.11776174604892731, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.0154, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.11634577810764313, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0158, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.11508087068796158, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0165, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.12009026855230331, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0141, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.15844415128231049, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.014, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.11803163588047028, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.014, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.11727108061313629, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.0157, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.09923852235078812, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0153, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.14936032891273499, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0174, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.10241764038801193, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0164, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.11330186575651169, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.016, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.14860020577907562, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0152, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.1636815220117569, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.015, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.11093862354755402, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0142, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.11333172768354416, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0141, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.10221018642187119, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0155, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.11085604131221771, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0158, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.17324623465538025, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0153, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.10631918907165527, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.0136, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.09826772660017014, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0148, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.1288551241159439, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11162615567445755, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0148, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.1307016909122467, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.017, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11215298622846603, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0154, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.09471828490495682, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0145, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10456835478544235, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.0147, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.08179501444101334, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0129, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.11810965836048126, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0151, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.0977158397436142, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0139, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.09650871902704239, + "learning_rate": 1.476067039948277e-05, + "loss": 0.0138, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.10262744128704071, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0145, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11392423510551453, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0147, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11900563538074493, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0163, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.10666237771511078, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0162, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.10767115652561188, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0152, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.1232513040304184, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0144, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.12271980196237564, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0167, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.12620051205158234, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0176, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11799243837594986, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0153, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.11633779853582382, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0154, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.10304611176252365, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0173, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.11186012625694275, + "learning_rate": 1.471107401829042e-05, + "loss": 0.0143, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.12054259330034256, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0174, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.1436678171157837, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0146, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.12709002196788788, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0171, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.11617328971624374, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0146, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.1117766946554184, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0178, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.12944459915161133, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0155, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11261707544326782, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0144, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.11154191195964813, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.013, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.11878378689289093, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0142, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.12532871961593628, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0176, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.10492552816867828, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0165, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.10776105523109436, + "learning_rate": 1.466148079674631e-05, + "loss": 0.016, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.10712877660989761, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0165, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.091713547706604, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0155, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.10679589211940765, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0185, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.11124882102012634, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0141, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.10930055379867554, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0148, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.10555030405521393, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0152, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.0836755633354187, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0153, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.11849679052829742, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0158, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.13186359405517578, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.0167, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.11089541018009186, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0189, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.09473541378974915, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0155, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.1151607483625412, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0138, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.10416971892118454, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0138, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.10269354283809662, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0158, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11249537020921707, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0152, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.14352446794509888, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0169, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10745783895254135, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0159, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.09522922337055206, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0164, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.13701501488685608, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0146, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.08020906895399094, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0139, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09878123551607132, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.015, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.09399155527353287, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.015, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.11354538053274155, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0175, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.09135859459638596, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0156, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.08270548284053802, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.014, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.11252828687429428, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0157, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.0985439196228981, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0163, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.12125496566295624, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0146, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.09888023883104324, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0176, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.10250411182641983, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.10701503604650497, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0151, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11525820940732956, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0157, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.09755205363035202, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0159, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09472718834877014, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.015, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10360235720872879, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.0149, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.09946251660585403, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0164, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.12811636924743652, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0135, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.127360999584198, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.016, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.12480203062295914, + "learning_rate": 1.45003312024933e-05, + "loss": 0.0142, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.10674259066581726, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0156, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.12259232252836227, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0149, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.08998828381299973, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0149, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.10316051542758942, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0145, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.14545974135398865, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0158, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.1868322491645813, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.015, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.12345045059919357, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0154, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.11794525384902954, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0177, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.12257082760334015, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0152, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.12155821174383163, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0159, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12235453724861145, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0147, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.1056676059961319, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.015, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.1309097409248352, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0145, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.11743862181901932, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0152, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.12476198375225067, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.016, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.08397012203931808, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0162, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12207506597042084, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0159, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.09591932594776154, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0137, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.12573577463626862, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0164, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1315455585718155, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0164, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.1468929648399353, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0165, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.15008211135864258, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0148, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.09235277771949768, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0153, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.1275569051504135, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0163, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.08151543140411377, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0129, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10390543937683105, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.015, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.11043227463960648, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0141, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.1271016001701355, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0163, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09329107403755188, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0134, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.10737643390893936, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0135, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12903659045696259, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0159, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.11276338994503021, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0147, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.14026683568954468, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0163, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.12503503262996674, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0172, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.10071003437042236, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0138, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10627497732639313, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0138, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.0770413801074028, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0135, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.09400150179862976, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0154, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.1549134999513626, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.015, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.09969423711299896, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.0135, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.09770094603300095, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0149, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.0848713293671608, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0128, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08489260822534561, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0142, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.14606203138828278, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.0144, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.1283080279827118, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0147, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.09238846600055695, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0154, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.09167003631591797, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0153, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.1081148013472557, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0142, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.19600129127502441, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0158, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.08589469641447067, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0148, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10786846280097961, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0123, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.12004593759775162, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0154, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10496815294027328, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0133, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.1324186772108078, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0152, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.10160262882709503, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0137, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.14738032221794128, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0167, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.1030116155743599, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0146, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.10793277621269226, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0156, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.11900396645069122, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0165, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.12115667760372162, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0144, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.12846702337265015, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0163, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.09553872793912888, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.015, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.12807151675224304, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0161, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.12759897112846375, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.017, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.10901748389005661, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0151, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.09706093370914459, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0129, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10346078872680664, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.0147, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.12170220911502838, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0136, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.24494236707687378, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0139, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.10071292519569397, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0131, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.13647668063640594, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0151, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09296777099370956, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.0147, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.1015978679060936, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.0147, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.09014145284891129, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0134, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.08798721432685852, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0136, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.08899430930614471, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.014, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.10661277174949646, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.0189, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.10300924628973007, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.0165, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.10107792913913727, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0152, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.12540587782859802, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0147, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.14788442850112915, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0154, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.1628517359495163, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0155, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.12338890135288239, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0156, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09656467288732529, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0132, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.15762092173099518, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0142, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.1834212690591812, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0156, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.091162770986557, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.0142, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.14656929671764374, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0163, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.08447254449129105, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0144, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10864629596471786, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0137, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.12742409110069275, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0134, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.09852556139230728, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.13655197620391846, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0143, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.09299396723508835, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0134, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.08303989470005035, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0136, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.09860356897115707, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.016, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.08682689070701599, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0134, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.09411480277776718, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.0149, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.10258188843727112, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.016, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.09782259166240692, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0165, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.10300199687480927, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0148, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.09208223968744278, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0175, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.09144087135791779, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0148, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.09090441465377808, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0153, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.10571702569723129, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0168, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.1070566177368164, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0154, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.12103649973869324, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.015, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.10678276419639587, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0165, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11873649060726166, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0154, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.11791808903217316, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0145, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.0986858382821083, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.016, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09249810129404068, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0127, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12782645225524902, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0154, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.10502453148365021, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0162, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.09005366265773773, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.0143, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11655479669570923, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0158, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.10148292779922485, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0155, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.11018893122673035, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0147, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.11230522394180298, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0148, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.09911413490772247, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0159, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.09358150511980057, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0154, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.12689724564552307, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0145, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.10737589001655579, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.013, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.12670603394508362, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0148, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.1459159404039383, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0162, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.14686602354049683, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0139, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10709210485219955, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0161, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.08311928808689117, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.015, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12497760355472565, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0161, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.08203094452619553, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0129, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.12578807771205902, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0145, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10697296261787415, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0147, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.10980327427387238, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0154, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.09966560453176498, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0126, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.09740720689296722, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0126, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.1452278196811676, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.013, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.10068526864051819, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0152, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.10273153334856033, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0134, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.1101747378706932, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0148, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.11141166090965271, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0133, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1535777896642685, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0155, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.08465548604726791, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0161, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.07614064961671829, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0149, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09852079302072525, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0135, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.13796597719192505, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0163, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1669035404920578, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0143, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09881193935871124, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0142, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.12325254082679749, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0132, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.16859179735183716, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0168, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.09475059062242508, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.014, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.11981360614299774, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0186, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.10775262862443924, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0151, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.10053553432226181, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0129, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.10618098825216293, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0154, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.12225139886140823, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0171, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.07435912638902664, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0152, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.10211736708879471, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0177, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12564173340797424, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.015, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.10412495583295822, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0157, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.12016817182302475, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0146, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.10957430303096771, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0141, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08035340160131454, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0146, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.1221131756901741, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0153, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.10435835272073746, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0133, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.11107036471366882, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0137, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.0575222373008728, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0149, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.076436348259449, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0159, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.09325376152992249, + "learning_rate": 1.380704261284659e-05, + "loss": 0.0163, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.11715361475944519, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0161, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.08567722141742706, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0149, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.12162303924560547, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.017, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.11919409781694412, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.0154, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.10694293677806854, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0138, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.10736183077096939, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0133, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11603212356567383, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0127, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.1180153489112854, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.0161, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.11968789249658585, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.0151, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.1250530481338501, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0136, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10276933014392853, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0174, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.14728401601314545, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.0138, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.09268351644277573, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0129, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12262973934412003, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0147, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.11068148910999298, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0157, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.09095525741577148, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0149, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.12430994212627411, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0154, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1246076300740242, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0153, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.08052714169025421, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.0144, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.1667853593826294, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0164, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.10500273108482361, + "learning_rate": 1.372053087451907e-05, + "loss": 0.013, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.10832768678665161, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0171, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.11245214939117432, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0158, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.14754413068294525, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0146, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.16368411481380463, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0175, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.1533215045928955, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0138, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.11878912895917892, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0134, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.10748414695262909, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0155, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.14132313430309296, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0139, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.09151305258274078, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0133, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10014014691114426, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0143, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.12393905967473984, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0154, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13815879821777344, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0155, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.11488667875528336, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.014, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.08504864573478699, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0149, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.12521564960479736, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0161, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.09525968134403229, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0137, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.12103450298309326, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.017, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.10642504692077637, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0139, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.15175847709178925, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0153, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11594653874635696, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.017, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.12788720428943634, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0171, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.12372581660747528, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0157, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.09206387400627136, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0155, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.09841199964284897, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0162, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.135525643825531, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0148, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.0982549637556076, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.0148, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.07432328909635544, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0124, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09830821305513382, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0139, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.10874123871326447, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0149, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.08979356288909912, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0189, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.08131087571382523, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.0128, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.1204582154750824, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0141, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11884505301713943, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0144, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.16451388597488403, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0134, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.11953870207071304, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0135, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12111926078796387, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0159, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.1181047335267067, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0142, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.12142453342676163, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0156, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09890097379684448, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.0137, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.13294243812561035, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0132, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.1397709846496582, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0161, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.10867296159267426, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.014, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11157155781984329, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.015, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09516428411006927, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0156, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.21431738138198853, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0158, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.09854292124509811, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0159, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.21810708940029144, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0152, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.14076393842697144, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0162, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.0892275720834732, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0154, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.09736334532499313, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0127, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.13880307972431183, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.015, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1091906875371933, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0147, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11605603992938995, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0168, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.11140985041856766, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0162, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.09057680517435074, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0161, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.12626419961452484, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0161, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.136195570230484, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0142, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.12415066361427307, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0121, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.12581513822078705, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.0151, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.11835133284330368, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0147, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.12255530059337616, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.0135, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10576335340738297, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0149, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.12575241923332214, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.014, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09812266379594803, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0152, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.11845924705266953, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0143, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.10975389182567596, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0147, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.09649760276079178, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0131, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.12912778556346893, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.0152, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10225717723369598, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0143, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.1712195724248886, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0144, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.1477300226688385, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0155, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.11692417412996292, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0146, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.1232081949710846, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0152, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.14507624506950378, + "learning_rate": 1.341604094324139e-05, + "loss": 0.017, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.13909533619880676, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0136, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.1438206136226654, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.0167, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.09637414664030075, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0134, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.15815401077270508, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.015, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.12961018085479736, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0159, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.11171260476112366, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0133, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.0907273143529892, + "learning_rate": 1.33872699666556e-05, + "loss": 0.015, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.11181363463401794, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0162, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.11502614617347717, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0169, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.14428892731666565, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0158, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.1440858244895935, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0134, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.07517627626657486, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0129, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.11884543299674988, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0147, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.08753574639558792, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0178, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.09831796586513519, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.014, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.12262003868818283, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.016, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.11444123834371567, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0142, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11559469252824783, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0138, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.09936029464006424, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0116, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.12810789048671722, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.0137, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.07923603802919388, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0136, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.09709552675485611, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0148, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.0924353376030922, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0152, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.10444460809230804, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0136, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.09510108828544617, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.014, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.11014958471059799, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.014, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.13198788464069366, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0153, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.09857133030891418, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0128, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.15041643381118774, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0148, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.13857626914978027, + "learning_rate": 1.329277958632556e-05, + "loss": 0.018, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.13004948198795319, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0142, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.09864714741706848, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0143, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12682363390922546, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0177, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.09303848445415497, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0141, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.13559986650943756, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0153, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.10143767297267914, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0131, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.1287422627210617, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0142, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.12000516802072525, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.016, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.10434254258871078, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.014, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08141680061817169, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0138, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08781418204307556, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0138, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.1355503350496292, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.015, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.1243978887796402, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0143, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.11088231950998306, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0154, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12559033930301666, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0132, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.16149592399597168, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.09809687733650208, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.0142, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.12586510181427002, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0165, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.12403467297554016, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.14458157122135162, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.0135, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.09313284605741501, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0165, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.14337806403636932, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0144, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.12650802731513977, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0126, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.112495556473732, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.0154, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11324562877416611, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.0139, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.11720716208219528, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.0146, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.1261121928691864, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0143, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.1302812099456787, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0157, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.12074442952871323, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0172, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.11716746538877487, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0158, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.11565797030925751, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0144, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.11653099209070206, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0155, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.13103044033050537, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0152, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.10341408103704453, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0138, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.11098334193229675, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.0146, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.11139992624521255, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.10786942392587662, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0143, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.10139934718608856, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0154, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.12619639933109283, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0153, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.13168810307979584, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0154, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.09545969218015671, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0151, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.09984135627746582, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0129, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10903751850128174, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0135, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.11882342398166656, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0151, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.1105988621711731, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0137, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.09874245524406433, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0142, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.10673530399799347, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0157, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.10264459252357483, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0162, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10769633203744888, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0133, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.16478870809078217, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0149, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.12630558013916016, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0128, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09824837744235992, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0144, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12136032432317734, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0124, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.10943865031003952, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0154, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.11394404619932175, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.0117, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.10907739400863647, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0144, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10671219974756241, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0156, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.10080806165933609, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0141, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.17395640909671783, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0145, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.12620997428894043, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0186, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.13434381783008575, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0149, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.09921885281801224, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.0149, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.13742485642433167, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0155, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.0998004674911499, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10389069467782974, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0138, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.11996421217918396, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.0147, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.09136862307786942, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0129, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.11111150681972504, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0159, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.11024610698223114, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0138, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.15376223623752594, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0154, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.13934020698070526, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0138, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.11208872497081757, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0131, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1513150930404663, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.10351566225290298, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0137, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11078765988349915, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0164, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.15622863173484802, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.0147, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.11603175848722458, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0145, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.1254235804080963, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0162, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.0900387316942215, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.015, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09852428734302521, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0172, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.13502047955989838, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.014, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.1460445076227188, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0156, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.10317493230104446, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0129, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.127616748213768, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0157, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.11560963094234467, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0142, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.11239839345216751, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0135, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.1503314971923828, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0157, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.1253356635570526, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0154, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.13217659294605255, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0127, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09087327867746353, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0144, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.09290582686662674, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0147, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.10273045301437378, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0165, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.14741986989974976, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.016, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.0804770216345787, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0125, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.12413053959608078, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0136, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.09822124987840652, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0176, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.09091541171073914, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0161, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.09487786144018173, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0148, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1009742021560669, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0147, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.08731451630592346, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0118, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.08123810589313507, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0134, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.10559827089309692, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0159, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.11313227564096451, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0164, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.11703694611787796, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0171, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.1083052009344101, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0168, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.13762544095516205, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0185, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.07979343086481094, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.09789150208234787, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0139, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.1905684620141983, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0158, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.1036764457821846, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0118, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.09413912892341614, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0137, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.2435447722673416, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.0132, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.09290791302919388, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0146, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.1211535856127739, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0151, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.1133081242442131, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.013, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.11362706869840622, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0171, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.10019481182098389, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0151, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.09489522129297256, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0143, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12912064790725708, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0138, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.11348122358322144, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0136, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.1325748860836029, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0156, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.11978204548358917, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0146, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.10022053122520447, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0137, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.07388417422771454, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0152, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.10844552516937256, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.0142, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.12381294369697571, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0132, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09625408798456192, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.0154, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.1207597553730011, + "learning_rate": 1.276827841111951e-05, + "loss": 0.016, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.12216818332672119, + "learning_rate": 1.276419082873553e-05, + "loss": 0.016, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.0969374030828476, + "learning_rate": 1.27601034161465e-05, + "loss": 0.0125, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.13043035566806793, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0145, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.10600738227367401, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0144, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.11113949120044708, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0135, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.09458228200674057, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0139, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10076881945133209, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0139, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1297597885131836, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0142, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.10059160739183426, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0142, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10716533660888672, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0149, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.1138077974319458, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0151, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.13693730533123016, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0136, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.08157949149608612, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0143, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.09702442586421967, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0121, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09029403328895569, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0133, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.1444508135318756, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0133, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.11026642471551895, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0161, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.11229082942008972, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0135, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.08556455373764038, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0166, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10684676468372345, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.0142, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.11970192193984985, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0132, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.10227229446172714, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.015, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.08301195502281189, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0153, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.1275268793106079, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.014, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.11613821983337402, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.015, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.10491939634084702, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0127, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.09623268246650696, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0136, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.111306332051754, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0137, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0863005742430687, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.0137, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.09434706717729568, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0142, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.10771634429693222, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0155, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.14157304167747498, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.0145, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11419932544231415, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0154, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.12308389693498611, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0133, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.10775505006313324, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.0144, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.10383253544569016, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0141, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.1012682244181633, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0144, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.12071908265352249, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0157, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.12535089254379272, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0149, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.08634085953235626, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0152, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.10035993158817291, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0138, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.13245077431201935, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0135, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.08713636547327042, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0144, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.10049428045749664, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0134, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.10432936251163483, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0163, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.16578327119350433, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0166, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.16106748580932617, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0125, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.11968962103128433, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0142, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1189485490322113, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0155, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.10795646905899048, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0139, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.08580315858125687, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0134, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.17969708144664764, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0159, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09675301611423492, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0146, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.10921840369701385, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0158, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.11278913170099258, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0133, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12176436185836792, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0138, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.1186685562133789, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0171, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.10296562314033508, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.015, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.15394653379917145, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0173, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.06815146654844284, + "learning_rate": 1.252333462101563e-05, + "loss": 0.014, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.1100328117609024, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0129, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.08578243851661682, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0145, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.11262256652116776, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.015, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.09644307941198349, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0183, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09589987248182297, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0145, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.09658264368772507, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0144, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.11963947862386703, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0118, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.14156748354434967, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0164, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.14387111365795135, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0151, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.11542942374944687, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0123, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.12162452191114426, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0165, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.13585466146469116, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0157, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.11272095143795013, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0159, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.1828201562166214, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.0162, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.09920861572027206, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0163, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.12618015706539154, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.016, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.13111864030361176, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0155, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.08486807346343994, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0126, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.14975349605083466, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0156, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.12495499104261398, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0154, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.11145244538784027, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0147, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.08242745697498322, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0149, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.10702498257160187, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0152, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.16176027059555054, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0145, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.0959012433886528, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0134, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10282868891954422, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0143, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.11747556924819946, + "learning_rate": 1.241332681912234e-05, + "loss": 0.0165, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10770756006240845, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0134, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.09925016760826111, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0126, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.09207499027252197, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.016, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.21546097099781036, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0131, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.10848663747310638, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0182, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.09382427483797073, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0134, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.10758832842111588, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0136, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.10075095295906067, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0154, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.09366318583488464, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0151, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.13190370798110962, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0154, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.1926490068435669, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0162, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.10018754750490189, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0127, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10250869393348694, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0146, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.08466067910194397, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0152, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.10397373139858246, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.0148, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14372816681861877, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0151, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11059669405221939, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0133, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09743087738752365, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0151, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.11265920102596283, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0145, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.08748266100883484, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0146, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.10545652359724045, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0133, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.09680107235908508, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0149, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.13201938569545746, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0152, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.12063595652580261, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0167, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.09986719489097595, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0153, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.09913598746061325, + "learning_rate": 1.230752865632506e-05, + "loss": 0.0151, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.1373804211616516, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0144, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.12088140845298767, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0137, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.08200784027576447, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0145, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.12470699846744537, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0154, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.10114862769842148, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.13169868290424347, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0159, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.08321736752986908, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0147, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.10108216106891632, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.012, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.08475375175476074, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0155, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.08946104347705841, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0145, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.07863342016935349, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0143, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.0906776711344719, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0123, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.11063490062952042, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0128, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.10001818835735321, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0133, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.12135739624500275, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0145, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.1233866810798645, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0133, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.09280609339475632, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0161, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.11677120625972748, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0144, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.11756281554698944, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.0163, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.08084359019994736, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.1415908932685852, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0149, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09107301384210587, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.013, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.11712655425071716, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.014, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.08366917073726654, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0125, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10391885787248611, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0146, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.1042654886841774, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0132, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.09504523873329163, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0148, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.12793844938278198, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0164, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.16210904717445374, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.15506179630756378, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.0153, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.11274553835391998, + "learning_rate": 1.218156580082751e-05, + "loss": 0.015, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.14852653443813324, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0172, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13558489084243774, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0156, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.08902081102132797, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0152, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12315832078456879, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0154, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.13528762757778168, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0145, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.1538013368844986, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0151, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.09894628077745438, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0154, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07696802169084549, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0136, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.09229687601327896, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0131, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.11709824949502945, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0146, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09195996820926666, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0142, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.11714771389961243, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.0157, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.11561911553144455, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0162, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.12904389202594757, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0129, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.10347782075405121, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0126, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.10054155439138412, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0126, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.1713176816701889, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0142, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10279975831508636, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0133, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.11513496935367584, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0137, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.096926249563694, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0123, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.1159687489271164, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.0145, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09745503962039948, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.014, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.15608133375644684, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.0144, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.08617804199457169, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.09285406023263931, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0123, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.11964505165815353, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0143, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1376206874847412, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0136, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.12900781631469727, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0148, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.10527875274419785, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0157, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.11285959929227829, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0135, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.09756818413734436, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0158, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.10695865005254745, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09360404312610626, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0129, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.100432388484478, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0126, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.10982918739318848, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0158, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.07767278701066971, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.015, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.10380839556455612, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0153, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.1225915253162384, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0167, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.0981304869055748, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0139, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.10055239498615265, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0146, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.09311894327402115, + "learning_rate": 1.201528720560666e-05, + "loss": 0.016, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.11859868466854095, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0163, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.15336444973945618, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0153, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.09175246208906174, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0125, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.0973501205444336, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0155, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.1643284410238266, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0142, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.1077687218785286, + "learning_rate": 1.19909852066355e-05, + "loss": 0.015, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.07722865790128708, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0125, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.149483323097229, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.015, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.11405838280916214, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0144, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.11132190376520157, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0142, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.10609457641839981, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0128, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.09458910673856735, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0134, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.10288971662521362, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0164, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.0970534011721611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0128, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.09571509063243866, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0129, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.08922260254621506, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0144, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0886709988117218, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.017, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.11732205003499985, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0141, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.11602790653705597, + "learning_rate": 1.193835918571248e-05, + "loss": 0.016, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.11643379926681519, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.0156, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.0927688404917717, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0149, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11533524841070175, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0133, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11001960933208466, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0135, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.12629783153533936, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.014, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.13336558640003204, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0121, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.11672402173280716, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.016, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.10014890134334564, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0136, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.0940401554107666, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.0151, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.1272585242986679, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0139, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.08492123335599899, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0151, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.1195707842707634, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0137, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.1095622330904007, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0167, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.09656022489070892, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0137, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1358073204755783, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0166, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.10981698334217072, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0141, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.07301078736782074, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0127, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.1040695309638977, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0143, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.16375352442264557, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.0136, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.19000792503356934, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0172, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11925465613603592, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0148, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.188803032040596, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.0151, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.09543585777282715, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0129, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11177625507116318, + "learning_rate": 1.184130718987356e-05, + "loss": 0.013, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.11703382432460785, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0153, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.12326977401971817, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0151, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.09018218517303467, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0147, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1641266793012619, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0146, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.08780542016029358, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0127, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09138940274715424, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0133, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.11768870800733566, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0128, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10298353433609009, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0141, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.10553838312625885, + "learning_rate": 1.180494808593063e-05, + "loss": 0.014, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10878341645002365, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0166, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10571706295013428, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0166, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.09349539875984192, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0126, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.10237022489309311, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0128, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.11938492953777313, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0167, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.10417849570512772, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0132, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.10322093963623047, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0141, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.09445950388908386, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.0146, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.1363389790058136, + "learning_rate": 1.176860863606944e-05, + "loss": 0.014, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.10975046455860138, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.013, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.12059775739908218, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0136, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.10529737174510956, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0158, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.11280884593725204, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0146, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.09665497392416, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0126, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.14059090614318848, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0142, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.10380115360021591, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.0132, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.10641331225633621, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0157, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.10845568776130676, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0135, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.10574200004339218, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0153, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.09128011763095856, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0142, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.10704760998487473, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0129, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.1180003359913826, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0141, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.11784005165100098, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0138, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.08203493803739548, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0144, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.10125461220741272, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.0142, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.12113944441080093, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0124, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.09266798943281174, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0156, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.11875846236944199, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0146, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.11041444540023804, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.013, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.13168254494667053, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0151, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.08208412677049637, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0132, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.08591622114181519, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0152, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.14605054259300232, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.014, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.12816093862056732, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0163, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.10751980543136597, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.013, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.1012972742319107, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0134, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13583813607692719, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0121, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.13573803007602692, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0143, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11719135195016861, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.0163, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.11966057121753693, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0133, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.11172359436750412, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0146, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10357552021741867, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0134, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.09571186453104019, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0143, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.13586629927158356, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0162, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.10461284965276718, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0134, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.182363823056221, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.014, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.12570780515670776, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0161, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.10319641977548599, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0136, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1066908910870552, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0123, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.10409434139728546, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0143, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.08230862766504288, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0128, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.110751211643219, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0148, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.09656667709350586, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0137, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.10626243054866791, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0127, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10294152796268463, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0151, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.09571471065282822, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0136, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.1162625253200531, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0129, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.09755981713533401, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0151, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.08586591482162476, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.016, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.11462593078613281, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.013, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.11438485980033875, + "learning_rate": 1.155904354214324e-05, + "loss": 0.016, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.15895701944828033, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0161, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.1135995015501976, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0128, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.11993709206581116, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0142, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.14105743169784546, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0118, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.08521834015846252, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0136, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.17094957828521729, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0159, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.1379796713590622, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.012, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.10417291522026062, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0138, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.09216596931219101, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0151, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.08398481458425522, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0124, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.1291581243276596, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0152, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.09349840134382248, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0127, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.08282706141471863, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0143, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.10378136485815048, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0146, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1504564881324768, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0144, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.0940636470913887, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0147, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.11142407357692719, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0127, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.16776388883590698, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0158, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.10488522052764893, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0132, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.09612040966749191, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0119, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.09318249672651291, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0116, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.08647014945745468, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0124, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.1067274957895279, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0136, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.08162420988082886, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0136, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.08949539810419083, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0136, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.11327249556779861, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0158, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.10963857173919678, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0141, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.10124490410089493, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.014, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10530558228492737, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0135, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.1251840591430664, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0127, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.14418582618236542, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0156, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.09544084221124649, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0137, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.09794268012046814, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0151, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.10741079598665237, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0127, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.10632490366697311, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0117, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08640208840370178, + "learning_rate": 1.141437135633602e-05, + "loss": 0.0119, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.09476958960294724, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0134, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.1141064390540123, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0147, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.1008499413728714, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0119, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.0853564441204071, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.12970876693725586, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0128, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11144055426120758, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0137, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.09443596750497818, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0127, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.19357536733150482, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0173, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.11861733347177505, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0145, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.13152548670768738, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0145, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11607249826192856, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0128, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.09456845372915268, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0145, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.12110379338264465, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.014, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.07872474938631058, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0128, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.1042177751660347, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0136, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.15168064832687378, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0126, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.09023087471723557, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0125, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.1110723614692688, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0121, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.09889500588178635, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0136, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.13798406720161438, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0151, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.14507459104061127, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0175, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10034530609846115, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0143, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.09978580474853516, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0142, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.11770957708358765, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0145, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.10433546453714371, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0148, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.1006259173154831, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.016, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.09567620605230331, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0133, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08685444295406342, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0116, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.10901349037885666, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.0137, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.15114818513393402, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0147, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.08353722840547562, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0146, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.14387859404087067, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0135, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.12125595659017563, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0148, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.11398079991340637, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0142, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.14072677493095398, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0149, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.11925073713064194, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0153, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.10301142185926437, + "learning_rate": 1.126604838049735e-05, + "loss": 0.013, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.11580533534288406, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.012, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.10776577144861221, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0143, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.1724127233028412, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0146, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.15091492235660553, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0128, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.10073890537023544, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0131, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11136317253112793, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0127, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.11012828350067139, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0133, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.12291619181632996, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0148, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.15791606903076172, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0144, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.12378063052892685, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.0137, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12597671151161194, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0172, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.1783512681722641, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0169, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.07375630736351013, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0139, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.07327539473772049, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0125, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.09731314331293106, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0144, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.0931868702173233, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0124, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.10977564752101898, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0132, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.0881611704826355, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0168, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.1239505335688591, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0133, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.10325739532709122, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0132, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11499917507171631, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0119, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.10131840407848358, + "learning_rate": 1.117803938925701e-05, + "loss": 0.014, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.10435229539871216, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0119, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.10805722326040268, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0136, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0759270042181015, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0167, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.1406354457139969, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0155, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1452491581439972, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.10772743821144104, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0148, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09850231558084488, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0155, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.08501581847667694, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0155, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11223053187131882, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0146, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.07870364189147949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0136, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.11768847703933716, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.0153, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.096342071890831, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0137, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.10153007507324219, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0139, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.0837850570678711, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0117, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.11286687850952148, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0148, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.14671675860881805, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0117, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.09201597422361374, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0161, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.11397483944892883, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0127, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.08533219248056412, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0126, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.18251192569732666, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0134, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.17834120988845825, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.014, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11550542712211609, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0127, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11998283863067627, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0149, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11872667074203491, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0155, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.10344526916742325, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.015, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.11288250237703323, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0149, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.16417740285396576, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0128, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09344452619552612, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0131, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11792241781949997, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0159, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10204951465129852, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0143, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.13064894080162048, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0151, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.08712288737297058, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0148, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.11878253519535065, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0123, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0907600000500679, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0135, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.08273347467184067, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0143, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.08222810924053192, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0112, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.11497339606285095, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0166, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.115402452647686, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0123, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.09611446410417557, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0128, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.12478714436292648, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0133, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.1713663637638092, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0118, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.13887059688568115, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0144, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.08571378141641617, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0154, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.1155383437871933, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0172, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.1123170331120491, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0153, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.13517709076404572, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0142, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.10638624429702759, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0121, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.11200621724128723, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0129, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10821404308080673, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.0128, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.10840310156345367, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0119, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.10352504998445511, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0144, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.10084252059459686, + "learning_rate": 1.097058241668254e-05, + "loss": 0.0128, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.11881101131439209, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.09931256622076035, + "learning_rate": 1.09626195543497e-05, + "loss": 0.0148, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.1260150521993637, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0125, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14202995598316193, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0133, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.15601785480976105, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0158, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.10701733827590942, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0139, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10180657356977463, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0138, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.0962422713637352, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.0124, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.13017651438713074, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0153, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.1295875608921051, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0154, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.10589175671339035, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0144, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.10173171758651733, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0142, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11211177706718445, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.0144, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.09875764697790146, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.014, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.11233554035425186, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0135, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.11327439546585083, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0136, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.10059859603643417, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0143, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.1160704717040062, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0132, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.15698356926441193, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0163, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.08235745877027512, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0134, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.12417344003915787, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0145, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.25982916355133057, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0155, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.12248262017965317, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0139, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.11425210535526276, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.015, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.12354899197816849, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0187, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.1102118268609047, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0137, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11539250612258911, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.015, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.12405017763376236, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0136, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.10405998677015305, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.014, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.0918804481625557, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0162, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.10429271310567856, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0128, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.09917982667684555, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0133, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.15390615165233612, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0133, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.11222631484270096, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0151, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.14449205994606018, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0165, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1332763433456421, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.015, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09963493794202805, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0132, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.12338289618492126, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0122, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.09369654208421707, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0127, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.13490226864814758, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0149, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.08246663957834244, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0158, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.0892438217997551, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0124, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.1275915652513504, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0135, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.150055930018425, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0118, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.08317377418279648, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0141, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.118314728140831, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.0131, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.086687833070755, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0128, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.10310143977403641, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0125, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.06559336930513382, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0135, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.11680027842521667, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0135, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.1381155252456665, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.015, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.08212172240018845, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0139, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.11066312342882156, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0145, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.11565965414047241, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0131, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.08814681321382523, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0122, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.11116618663072586, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0156, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.10491067916154861, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0157, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.15800951421260834, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0141, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.10169819742441177, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0156, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.07719215005636215, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0127, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.08958051353693008, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0145, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.1171223521232605, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0141, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.09662648290395737, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0121, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.09235699474811554, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0129, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.127174973487854, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0157, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.10298880189657211, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0146, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08706901967525482, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0132, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.10898775607347488, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0117, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.12178633362054825, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0129, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.10881572961807251, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0138, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.12326113879680634, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.013, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.13331058621406555, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0129, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.11921956390142441, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0154, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.10736659914255142, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.014, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09894116222858429, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0141, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.09651704877614975, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0146, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.11836998164653778, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.0176, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.1108580157160759, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0129, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.1281266212463379, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0138, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.13581599295139313, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0149, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10911770164966583, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0162, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.14131440222263336, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0126, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.10660075396299362, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0144, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.12547564506530762, + "learning_rate": 1.063722589416218e-05, + "loss": 0.0142, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11465640366077423, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0143, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.1274002492427826, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0145, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.10782597959041595, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0144, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.08273816108703613, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0126, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.12750636041164398, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0113, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.10962793231010437, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0132, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11728726327419281, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0146, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.10677216202020645, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0146, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.0974075049161911, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.0129, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.09190504997968674, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0129, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.10864423215389252, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0122, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.08833710849285126, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0141, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.14306439459323883, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.013, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.11269631236791611, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0115, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.08850698918104172, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0116, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10025947540998459, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0116, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.07444360852241516, + "learning_rate": 1.057003955600053e-05, + "loss": 0.013, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.10606181621551514, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0133, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.11394638568162918, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0143, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.12015067040920258, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0128, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.11465302109718323, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0117, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.09164674580097198, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0116, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.10532126575708389, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0128, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.16767282783985138, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0137, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.16740484535694122, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0151, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08568812906742096, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0135, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.24568453431129456, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0138, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.13317807018756866, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0142, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11310465633869171, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0139, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.09677092730998993, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0123, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.13912886381149292, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0152, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.0902378037571907, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0124, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09430436044931412, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0116, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.13361187279224396, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0161, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09588316082954407, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0135, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.13104380667209625, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0161, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.11476746946573257, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0137, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.087586909532547, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0124, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.12136878073215485, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0156, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.09384764730930328, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0137, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.10422758758068085, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0142, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.11391496658325195, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0125, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.12965570390224457, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0137, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.13042569160461426, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.0159, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.10433127731084824, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0129, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.0989370346069336, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0132, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.08399394899606705, + "learning_rate": 1.045171352558859e-05, + "loss": 0.013, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.11693604290485382, + "learning_rate": 1.044777463361479e-05, + "loss": 0.0148, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.12304933369159698, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0121, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.10788783431053162, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0138, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11043704301118851, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0135, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.11343314498662949, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0158, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.10075241327285767, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0143, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.08752909302711487, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.012, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.11036614328622818, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0126, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.09241441637277603, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.0157, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.1369611620903015, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.0132, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.12782007455825806, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0136, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.140136256814003, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0144, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11508572846651077, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.0136, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.14334864914417267, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0138, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.10154326260089874, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0135, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.11290924996137619, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0134, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.13472969830036163, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0144, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.1215982660651207, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0125, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.10628669708967209, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0125, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.07950089126825333, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0135, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.14542941749095917, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0145, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.1189931258559227, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0115, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.11141669005155563, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0135, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.08776746690273285, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0136, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10942544043064117, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0139, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.15493649244308472, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0123, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.1485723853111267, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.0141, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.0886952131986618, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0124, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.09021062403917313, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0132, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.17778685688972473, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0126, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.11892827600240707, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0125, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.14066243171691895, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0135, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.10870271921157837, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0147, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.10636188089847565, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0156, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.11178954690694809, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.014, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.09059949964284897, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.0127, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.11947090923786163, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0145, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.1133369505405426, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.014, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.12451290339231491, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0139, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10219574719667435, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0138, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.09843198210000992, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0144, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.09784907102584839, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0155, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1259097307920456, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0156, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.08604080229997635, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0113, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.11517608165740967, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0131, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.09721313416957855, + "learning_rate": 1.027088684228894e-05, + "loss": 0.011, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.08404546976089478, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0164, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.11207510530948639, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0151, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.1101691946387291, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0129, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.2047414630651474, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0143, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.08887176960706711, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0139, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.08608511090278625, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0162, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10840177536010742, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0133, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.1228431761264801, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.0173, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.09171874076128006, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0119, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.1357734203338623, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0151, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.13374143838882446, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0134, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09826765209436417, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0112, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.30271241068840027, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0128, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.1474897861480713, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.0142, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.11385788768529892, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0121, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.1145087331533432, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0125, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.15277381241321564, + "learning_rate": 1.020425036775911e-05, + "loss": 0.011, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10591624677181244, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0123, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.09938671439886093, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0136, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.12457235157489777, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0145, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.10550948232412338, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0127, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.12241430580615997, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.013, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.12996822595596313, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0127, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.10886852443218231, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0124, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.10862760245800018, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0124, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.12071339040994644, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0152, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.12002567946910858, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0144, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.0731884092092514, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.014, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.09087152034044266, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0122, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.14147141575813293, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0126, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.10024842619895935, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.013, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.12605364620685577, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0129, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.07398458570241928, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.012, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.09070646017789841, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0139, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.06930165737867355, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.012, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.10134197026491165, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0131, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.12521374225616455, + "learning_rate": 1.012598938117799e-05, + "loss": 0.012, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.10238701850175858, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0139, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.10689418017864227, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0138, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.0954018235206604, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0125, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.1394306868314743, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0125, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.11217380315065384, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0143, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.1513020396232605, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.0122, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.11155983060598373, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0132, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.1257336288690567, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.015, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.10818599164485931, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0163, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.15947701036930084, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0142, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.10929335653781891, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0129, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.08994319289922714, + "learning_rate": 1.007910370568197e-05, + "loss": 0.011, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.09995368868112564, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0131, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.10167150944471359, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.0142, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1480625420808792, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0129, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.11566069722175598, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.013, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.10494371503591537, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0132, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.09531164914369583, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0116, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.0887216180562973, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0133, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.09848874062299728, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0122, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.14705902338027954, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0135, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.1412736475467682, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11742296814918518, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0143, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.11000296473503113, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0144, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.111237533390522, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0125, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.1085008904337883, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0133, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.12815521657466888, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0138, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.11128585785627365, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.12439922988414764, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0139, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.11863227933645248, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0154, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.13517265021800995, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0128, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.0960967019200325, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0123, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.0953422263264656, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0121, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.1965440809726715, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.0924379900097847, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0112, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.08345641940832138, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0146, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.10248678177595139, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0121, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.2980184257030487, + "learning_rate": 9.977703371038337e-06, + "loss": 0.013, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.11192528903484344, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0133, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.21329663693904877, + "learning_rate": 9.969913958481688e-06, + "loss": 0.0116, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.0922258049249649, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0159, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.12964527308940887, + "learning_rate": 9.962126073932371e-06, + "loss": 0.0143, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.11362354457378387, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0137, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.0902123972773552, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0126, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.08770131319761276, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0123, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.09604862332344055, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0131, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.08182749897241592, + "learning_rate": 9.942663063115248e-06, + "loss": 0.013, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.17937488853931427, + "learning_rate": 9.938771611983594e-06, + "loss": 0.0158, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.1343632936477661, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0139, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.14316903054714203, + "learning_rate": 9.93098986311647e-06, + "loss": 0.012, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11366818100214005, + "learning_rate": 9.927099565971976e-06, + "loss": 0.0129, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0949450209736824, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0144, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.13775621354579926, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0134, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.10221140086650848, + "learning_rate": 9.915430987239967e-06, + "loss": 0.012, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.11101876199245453, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0121, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.12785188853740692, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0146, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.1400655061006546, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0115, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.09468253701925278, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0122, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.08724251389503479, + "learning_rate": 9.895991082746168e-06, + "loss": 0.0143, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.10251675546169281, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.11980220675468445, + "learning_rate": 9.888217832192542e-06, + "loss": 0.015, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.12473925948143005, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0144, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.08260339498519897, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0122, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.11015481501817703, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0154, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1120007336139679, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0143, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09239251911640167, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0146, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10413528233766556, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0144, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.0863887220621109, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0143, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.0963623896241188, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0129, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.2090010792016983, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0155, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12460873275995255, + "learning_rate": 9.84937491900087e-06, + "loss": 0.012, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.11049462109804153, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0134, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.15180432796478271, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0142, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.13392038643360138, + "learning_rate": 9.837729659166156e-06, + "loss": 0.013, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.11534144729375839, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0134, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.1077096238732338, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0135, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.10903865098953247, + "learning_rate": 9.826087927694543e-06, + "loss": 0.014, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.09442178905010223, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0132, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.12171642482280731, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0143, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.08623645454645157, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0135, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.10249772667884827, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0151, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.16136185824871063, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0113, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.15169787406921387, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0161, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.095759317278862, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0143, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.11120729893445969, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0127, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09804334491491318, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0123, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.11462689191102982, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0111, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10223866254091263, + "learning_rate": 9.783431894635035e-06, + "loss": 0.012, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.1001039668917656, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0149, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.13274405896663666, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0136, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.09141503274440765, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0135, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.1291927844285965, + "learning_rate": 9.767932474193758e-06, + "loss": 0.0136, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10977526009082794, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0135, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.1167440190911293, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0138, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.08445189893245697, + "learning_rate": 9.756312079913248e-06, + "loss": 0.0129, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.2519867718219757, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0146, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.08974999934434891, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0152, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.1252872496843338, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0147, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.16116158664226532, + "learning_rate": 9.740823797373594e-06, + "loss": 0.013, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.08403292298316956, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0119, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.11238662898540497, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0154, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10106278210878372, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0136, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.08145973831415176, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0131, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.15683840215206146, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0132, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.14468176662921906, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0146, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.08105593919754028, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0122, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.12362894415855408, + "learning_rate": 9.709866422291219e-06, + "loss": 0.013, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.09933894127607346, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0156, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.09758161753416061, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0145, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.1307433545589447, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0139, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.1336813122034073, + "learning_rate": 9.694397367364635e-06, + "loss": 0.015, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.08841264992952347, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0162, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09142819046974182, + "learning_rate": 9.68666525627968e-06, + "loss": 0.0132, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11312894523143768, + "learning_rate": 9.682799805859474e-06, + "loss": 0.0129, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.14984415471553802, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0147, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.104182668030262, + "learning_rate": 9.675070116731382e-06, + "loss": 0.013, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.10078790783882141, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0131, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.08538635075092316, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0135, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.11059115827083588, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0134, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.10521579533815384, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0131, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.1130848377943039, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0137, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.09023619443178177, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0125, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.11787812411785126, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0136, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.14386765658855438, + "learning_rate": 9.6441675594234e-06, + "loss": 0.0153, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.10471580922603607, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0128, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.1340251863002777, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0139, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.1131536215543747, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0128, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10629499703645706, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0147, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.10359197854995728, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0121, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.11530483514070511, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0136, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.09336108714342117, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0124, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.08179444819688797, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0124, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.12341088801622391, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0125, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.12123409658670425, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0149, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.11406739801168442, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0141, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.07922708988189697, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0114, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.09156256914138794, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0121, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.09705155342817307, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0134, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.10157317668199539, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0157, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.09271441400051117, + "learning_rate": 9.582440688952473e-06, + "loss": 0.0113, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.09375326335430145, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0109, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.07997307926416397, + "learning_rate": 9.57473221124502e-06, + "loss": 0.0137, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.11743918806314468, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0138, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.09150344878435135, + "learning_rate": 9.567025381590669e-06, + "loss": 0.013, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.09968695789575577, + "learning_rate": 9.563172585515032e-06, + "loss": 0.011, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.10516709089279175, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0114, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.09716377407312393, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0132, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.12646295130252838, + "learning_rate": 9.551616675805319e-06, + "loss": 0.0145, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.11036434769630432, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0153, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10279295593500137, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0138, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.08990927040576935, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0132, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.10534624755382538, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0138, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.09507470577955246, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0114, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.101425401866436, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0137, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.11853020638227463, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0121, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.13193756341934204, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0128, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.10581472516059875, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0123, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.12424535304307938, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0139, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.1036083921790123, + "learning_rate": 9.509276927101462e-06, + "loss": 0.0126, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.07990803569555283, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0112, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.098060242831707, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0136, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.11892619729042053, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0126, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.10609585791826248, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0127, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.10905738174915314, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0123, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.0981525182723999, + "learning_rate": 9.486203760067013e-06, + "loss": 0.012, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.16955061256885529, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0119, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.10360468924045563, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0136, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.08537023514509201, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0112, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.1317787766456604, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0128, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10730036348104477, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0151, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09596534818410873, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0117, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.10839905589818954, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0138, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.12511657178401947, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0138, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.1078735962510109, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0127, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.10344097018241882, + "learning_rate": 9.447782011716307e-06, + "loss": 0.0119, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.18708902597427368, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0139, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.09599670022726059, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0125, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.13859155774116516, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0143, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09208878129720688, + "learning_rate": 9.432425102360072e-06, + "loss": 0.0126, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.21019718050956726, + "learning_rate": 9.428586931343613e-06, + "loss": 0.0117, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.09587983787059784, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0122, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.09626271575689316, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0124, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.1051478385925293, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0154, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.10166551172733307, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0119, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.09312118589878082, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0145, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.11722584068775177, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0126, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.0939759686589241, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0124, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.11434069275856018, + "learning_rate": 9.39789681972375e-06, + "loss": 0.012, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12582115828990936, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0137, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.07123387604951859, + "learning_rate": 9.390228540420876e-06, + "loss": 0.0155, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.08926647901535034, + "learning_rate": 9.386395039661539e-06, + "loss": 0.014, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.07371679693460464, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0128, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.1008603498339653, + "learning_rate": 9.378729317382715e-06, + "loss": 0.012, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.10988970100879669, + "learning_rate": 9.37489709644538e-06, + "loss": 0.0136, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.09886831045150757, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0123, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.10959886759519577, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0135, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.12443319708108902, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0161, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10207810997962952, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0146, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.11660879850387573, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0127, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.13432076573371887, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0142, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.11405819654464722, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0135, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.12952250242233276, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0119, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.09084811806678772, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0173, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.12252718955278397, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0099, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1293795108795166, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0144, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11287739127874374, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0164, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.1253224015235901, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0133, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.10046776384115219, + "learning_rate": 9.321290990574668e-06, + "loss": 0.014, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.08746109157800674, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0128, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.08306485414505005, + "learning_rate": 9.313639857014607e-06, + "loss": 0.0115, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.1340319961309433, + "learning_rate": 9.3098149378516e-06, + "loss": 0.0137, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.09587698429822922, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0143, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.14915677905082703, + "learning_rate": 9.302166396212059e-06, + "loss": 0.0143, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.09313366562128067, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0115, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.12605343759059906, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0146, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.12807077169418335, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0113, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.1186203733086586, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0119, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.10223432630300522, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0141, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.12712709605693817, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0132, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.12993593513965607, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0162, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.12273230403661728, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0133, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.1391134113073349, + "learning_rate": 9.26776941165236e-06, + "loss": 0.014, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.11695246398448944, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0137, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.1214752346277237, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0122, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.13961699604988098, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0117, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.1402655988931656, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0159, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.10683967918157578, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0146, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.08510883897542953, + "learning_rate": 9.244857665179213e-06, + "loss": 0.013, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.10308418422937393, + "learning_rate": 9.241040568803432e-06, + "loss": 0.0141, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.09736710786819458, + "learning_rate": 9.23722390978261e-06, + "loss": 0.0139, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.13270588219165802, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0118, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.08694430440664291, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0132, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.07487893849611282, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0123, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.1315941959619522, + "learning_rate": 9.221961653045449e-06, + "loss": 0.0115, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10982084274291992, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0146, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.09258668124675751, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.11609145998954773, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0143, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.1349419355392456, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0139, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.08282078802585602, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0108, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.12237592041492462, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0149, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.07864762842655182, + "learning_rate": 9.195269608854445e-06, + "loss": 0.012, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.1032051146030426, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0112, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.1346377581357956, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0132, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.08922939747571945, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0118, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.09849593788385391, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0141, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.07525688409805298, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0126, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.1287003457546234, + "learning_rate": 9.172407902473605e-06, + "loss": 0.0117, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.11016935855150223, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0123, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.09082796424627304, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0119, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.12462439388036728, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0133, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.10698465257883072, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0126, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.07600131630897522, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0133, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.09345563501119614, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0144, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.1186055988073349, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0151, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.12372780591249466, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0133, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.116783007979393, + "learning_rate": 9.138145235699377e-06, + "loss": 0.0139, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.08260510116815567, + "learning_rate": 9.134340494716546e-06, + "loss": 0.012, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.1815570890903473, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0125, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.12819825112819672, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0133, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.09238126873970032, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0115, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.09430833160877228, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0119, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.7733426094055176, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0128, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.10260160267353058, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0126, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.11146093159914017, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0136, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.11368793249130249, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0141, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.08438809961080551, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0134, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.11850261688232422, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.13420401513576508, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0128, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.09115619212388992, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0127, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.09974579513072968, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0141, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.10242673754692078, + "learning_rate": 9.081121025468816e-06, + "loss": 0.0133, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.08850608766078949, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0113, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.08285540342330933, + "learning_rate": 9.073525417675752e-06, + "loss": 0.0122, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.08347760140895844, + "learning_rate": 9.069728288749965e-06, + "loss": 0.011, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.12695780396461487, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0134, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.10132371634244919, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0143, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.07370374351739883, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0112, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.0993623211979866, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0138, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.07118473947048187, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0134, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.10804349929094315, + "learning_rate": 9.046954982950699e-06, + "loss": 0.014, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.12209299206733704, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0127, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.0896674394607544, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0134, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.10866492241621017, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0145, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.06327113509178162, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0125, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.1044771745800972, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0138, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10678671300411224, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0119, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.10146090388298035, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0122, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.11703896522521973, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0126, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.08116868138313293, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0117, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.16623066365718842, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0135, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.07943270355463028, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0128, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.17520980536937714, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0141, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.10311958938837051, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0124, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09508595615625381, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0139, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.11690590530633926, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0139, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.11585275083780289, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0119, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.08755052834749222, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0152, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.15034662187099457, + "learning_rate": 8.978732966762205e-06, + "loss": 0.013, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.11793015897274017, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0147, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.14666646718978882, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0159, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1565992683172226, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0152, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.11420148611068726, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0134, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.0945156067609787, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0105, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.0971534475684166, + "learning_rate": 8.95602513567626e-06, + "loss": 0.012, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.12953956425189972, + "learning_rate": 8.952242101985645e-06, + "loss": 0.0144, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.11144538223743439, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0138, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.11536341160535812, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0118, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.08281587064266205, + "learning_rate": 8.940895757786624e-06, + "loss": 0.0091, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.10917636007070541, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0116, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.11539111286401749, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0142, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.09897308051586151, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0141, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.12034150213003159, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0142, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.10506895184516907, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0145, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10428126901388168, + "learning_rate": 8.918215501162718e-06, + "loss": 0.013, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.0966438427567482, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0126, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.09688732028007507, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0136, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10319554805755615, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0133, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11589661985635757, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0136, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.08819710463285446, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0145, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.09449224174022675, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0141, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.11158958822488785, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0124, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.10327963531017303, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0138, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.070926733314991, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0143, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.08468566834926605, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0131, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11933926492929459, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0128, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10370025038719177, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0138, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.13985954225063324, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0144, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.08677119761705399, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0141, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.1023884266614914, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0113, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.14462946355342865, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0132, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.12994524836540222, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0144, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.08624777942895889, + "learning_rate": 8.850274743290876e-06, + "loss": 0.013, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.10459248721599579, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0117, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.12475694715976715, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0131, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09973360598087311, + "learning_rate": 8.838965950000907e-06, + "loss": 0.0142, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.10204580426216125, + "learning_rate": 8.835197287635673e-06, + "loss": 0.0136, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.08137936890125275, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0111, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.09507330507040024, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0128, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.11097851395606995, + "learning_rate": 8.823894110741103e-06, + "loss": 0.0145, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.12435153126716614, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0122, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.1186857596039772, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0123, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10177282989025116, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0119, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.08287174254655838, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0129, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.12975555658340454, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0138, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.07198113948106766, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0116, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.12089285254478455, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0126, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.09319751709699631, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0123, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.11510422080755234, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0147, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.1134580671787262, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0119, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12099597603082657, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0116, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.11170639842748642, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0111, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.0979531928896904, + "learning_rate": 8.774962556249487e-06, + "loss": 0.0137, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.0814109742641449, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0143, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.09910739213228226, + "learning_rate": 8.767441700331522e-06, + "loss": 0.0129, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.10014151781797409, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0125, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.14257559180259705, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.08456768840551376, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0138, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.19197505712509155, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0156, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.10880734771490097, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0143, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.12371200323104858, + "learning_rate": 8.744890501433196e-06, + "loss": 0.0117, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.08023890107870102, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0118, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.10231798142194748, + "learning_rate": 8.737377232363901e-06, + "loss": 0.0107, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.06945646554231644, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0131, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.09935662895441055, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0122, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.15378066897392273, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0142, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.08416267484426498, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0127, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.0862744078040123, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0132, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.11481238156557083, + "learning_rate": 8.71484884879911e-06, + "loss": 0.013, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.09441596269607544, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0126, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08091022074222565, + "learning_rate": 8.707343203095797e-06, + "loss": 0.0126, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.1480294167995453, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0121, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.12529131770133972, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0132, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.12804464995861053, + "learning_rate": 8.696088319396272e-06, + "loss": 0.013, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.10956594347953796, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0129, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.11884522438049316, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0127, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.09215931594371796, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0122, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.1411169320344925, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0139, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.07872501015663147, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0139, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.083979532122612, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0102, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.12711060047149658, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0148, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.10543761402368546, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0131, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.13125181198120117, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0142, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.13816513121128082, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0122, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.09294828027486801, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0126, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.10944795608520508, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0142, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.11008825153112411, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0122, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.10732167959213257, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0138, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.10374412685632706, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0136, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.1021609753370285, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0138, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.10304085910320282, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0125, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.08420020341873169, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0119, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.11911525577306747, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0126, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.07106399536132812, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0117, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.08642835915088654, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0125, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1176094189286232, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0146, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08892391622066498, + "learning_rate": 8.60994458849601e-06, + "loss": 0.012, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.12898282706737518, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0122, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.12732256948947906, + "learning_rate": 8.602465911610909e-06, + "loss": 0.0164, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.1723143309354782, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0146, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.10019178688526154, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0122, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.15125800669193268, + "learning_rate": 8.591251540871672e-06, + "loss": 0.0138, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.08618403971195221, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0106, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.12111509591341019, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0125, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.11654779314994812, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0133, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.12341999262571335, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0136, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.09898588061332703, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0121, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.09825480729341507, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0119, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.09357906877994537, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0125, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13970960676670074, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0131, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.14642572402954102, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0117, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.20111511647701263, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0122, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.12268098443746567, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0127, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.11043815314769745, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0109, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.12570181488990784, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0147, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.1037122830748558, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0134, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.07811714708805084, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0121, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.07688374072313309, + "learning_rate": 8.53151572917108e-06, + "loss": 0.013, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.12285313010215759, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0135, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11350435763597488, + "learning_rate": 8.524057571200438e-06, + "loss": 0.013, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.1195593997836113, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.12681595981121063, + "learning_rate": 8.516601380450119e-06, + "loss": 0.0116, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.10786957293748856, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0117, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.12382174283266068, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0123, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08464337885379791, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0139, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.11292237043380737, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0142, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.13033312559127808, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0125, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.13949616253376007, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0153, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.09846726804971695, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0115, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.12679991126060486, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0149, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.10708379000425339, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0145, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.0848502367734909, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0135, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.13396726548671722, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0142, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13248398900032043, + "learning_rate": 8.471905674372712e-06, + "loss": 0.014, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.10274169594049454, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0139, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.08507993817329407, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0112, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.10394402593374252, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0146, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.10293199867010117, + "learning_rate": 8.457022946052628e-06, + "loss": 0.015, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.16253890097141266, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0128, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.07673029601573944, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0125, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.1329779177904129, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0131, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.0754442885518074, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0125, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.09978003799915314, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0124, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.12813764810562134, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0144, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.08480092883110046, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0124, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.10068657249212265, + "learning_rate": 8.42728135848896e-06, + "loss": 0.012, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.08405598253011703, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0122, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.13025622069835663, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0117, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.11026641726493835, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0127, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.10400493443012238, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0138, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.10374867916107178, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0138, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.128962442278862, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0124, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.07732204347848892, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0124, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.14872954785823822, + "learning_rate": 8.39757171680725e-06, + "loss": 0.013, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.13792090117931366, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0175, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.07409467548131943, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0123, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.11308597028255463, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0149, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.1202300488948822, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0139, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.08481356501579285, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0124, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.07334130257368088, + "learning_rate": 8.375310536902511e-06, + "loss": 0.014, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.09884555637836456, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0137, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.16476134955883026, + "learning_rate": 8.36789416540753e-06, + "loss": 0.0142, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.12900881469249725, + "learning_rate": 8.364186735016254e-06, + "loss": 0.0138, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.10485891997814178, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0146, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.13036216795444489, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0111, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.13155606389045715, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0146, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.11586495488882065, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0135, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.1043778583407402, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0137, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.12916314601898193, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.1257125288248062, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0117, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.20916301012039185, + "learning_rate": 8.334545457592674e-06, + "loss": 0.014, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.13241995871067047, + "learning_rate": 8.330842572848802e-06, + "loss": 0.011, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.1633574515581131, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0141, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.08672885596752167, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0141, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.09803608059883118, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0143, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.13284195959568024, + "learning_rate": 8.316036101469748e-06, + "loss": 0.011, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10581238567829132, + "learning_rate": 8.312335751929666e-06, + "loss": 0.0144, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.11561308056116104, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0111, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.09897869825363159, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0116, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.129640132188797, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0142, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.08553853631019592, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.09117261320352554, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0134, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.1338709443807602, + "learning_rate": 8.290144330085352e-06, + "loss": 0.013, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.11192607134580612, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0122, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.0983581617474556, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0149, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.10853386670351028, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0145, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.09507740288972855, + "learning_rate": 8.275360236495652e-06, + "loss": 0.011, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.08634619414806366, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0146, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.12652689218521118, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0145, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.13687895238399506, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0115, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.11267609149217606, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0139, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.0999043732881546, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0125, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.09523805975914001, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0137, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09847134351730347, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0153, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.0827624723315239, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0093, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.09830755740404129, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0119, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.14614145457744598, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0149, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.1254947930574417, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0129, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.11176569014787674, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0147, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.06597425043582916, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0132, + "step": 80000 + }, + { + "epoch": 23.279022403258654, + "grad_norm": 0.19944901764392853, + "learning_rate": 8.223680373918572e-06, + "loss": 0.0147, + "step": 80010 + }, + { + "epoch": 23.2819319173698, + "grad_norm": 0.09820935130119324, + "learning_rate": 8.219992805682155e-06, + "loss": 0.013, + "step": 80020 + }, + { + "epoch": 23.28484143148094, + "grad_norm": 0.07166675478219986, + "learning_rate": 8.216305752342533e-06, + "loss": 0.0135, + "step": 80030 + }, + { + "epoch": 23.287750945592087, + "grad_norm": 0.08634001016616821, + "learning_rate": 8.21261921417972e-06, + "loss": 0.0128, + "step": 80040 + }, + { + "epoch": 23.29066045970323, + "grad_norm": 0.1467038244009018, + "learning_rate": 8.20893319147369e-06, + "loss": 0.0131, + "step": 80050 + }, + { + "epoch": 23.293569973814375, + "grad_norm": 0.0938076600432396, + "learning_rate": 8.205247684504364e-06, + "loss": 0.0122, + "step": 80060 + }, + { + "epoch": 23.296479487925517, + "grad_norm": 0.06792404502630234, + "learning_rate": 8.20156269355164e-06, + "loss": 0.012, + "step": 80070 + }, + { + "epoch": 23.29938900203666, + "grad_norm": 0.06738271564245224, + "learning_rate": 8.197878218895356e-06, + "loss": 0.0114, + "step": 80080 + }, + { + "epoch": 23.302298516147804, + "grad_norm": 0.16149401664733887, + "learning_rate": 8.194194260815338e-06, + "loss": 0.0121, + "step": 80090 + }, + { + "epoch": 23.305208030258946, + "grad_norm": 0.11988461762666702, + "learning_rate": 8.190510819591348e-06, + "loss": 0.0121, + "step": 80100 + }, + { + "epoch": 23.30811754437009, + "grad_norm": 0.09137140959501266, + "learning_rate": 8.186827895503117e-06, + "loss": 0.0125, + "step": 80110 + }, + { + "epoch": 23.311027058481233, + "grad_norm": 0.0880751833319664, + "learning_rate": 8.183145488830355e-06, + "loss": 0.0116, + "step": 80120 + }, + { + "epoch": 23.313936572592375, + "grad_norm": 0.09596525132656097, + "learning_rate": 8.179463599852705e-06, + "loss": 0.0117, + "step": 80130 + }, + { + "epoch": 23.31684608670352, + "grad_norm": 0.13169355690479279, + "learning_rate": 8.175782228849769e-06, + "loss": 0.0121, + "step": 80140 + }, + { + "epoch": 23.319755600814663, + "grad_norm": 0.1044575497508049, + "learning_rate": 8.17210137610115e-06, + "loss": 0.0132, + "step": 80150 + }, + { + "epoch": 23.322665114925808, + "grad_norm": 0.13589617609977722, + "learning_rate": 8.168421041886373e-06, + "loss": 0.0118, + "step": 80160 + }, + { + "epoch": 23.32557462903695, + "grad_norm": 0.10503333061933517, + "learning_rate": 8.164741226484925e-06, + "loss": 0.0108, + "step": 80170 + }, + { + "epoch": 23.328484143148096, + "grad_norm": 0.09190402179956436, + "learning_rate": 8.161061930176272e-06, + "loss": 0.0129, + "step": 80180 + }, + { + "epoch": 23.331393657259238, + "grad_norm": 0.104253388941288, + "learning_rate": 8.15738315323984e-06, + "loss": 0.0114, + "step": 80190 + }, + { + "epoch": 23.33430317137038, + "grad_norm": 0.10191919654607773, + "learning_rate": 8.153704895955e-06, + "loss": 0.0128, + "step": 80200 + }, + { + "epoch": 23.337212685481525, + "grad_norm": 0.0824175626039505, + "learning_rate": 8.150027158601084e-06, + "loss": 0.0126, + "step": 80210 + }, + { + "epoch": 23.340122199592667, + "grad_norm": 0.08966456353664398, + "learning_rate": 8.146349941457401e-06, + "loss": 0.014, + "step": 80220 + }, + { + "epoch": 23.343031713703812, + "grad_norm": 0.0856788232922554, + "learning_rate": 8.142673244803215e-06, + "loss": 0.01, + "step": 80230 + }, + { + "epoch": 23.345941227814954, + "grad_norm": 0.1027766764163971, + "learning_rate": 8.138997068917736e-06, + "loss": 0.0134, + "step": 80240 + }, + { + "epoch": 23.3488507419261, + "grad_norm": 0.11636431515216827, + "learning_rate": 8.13532141408015e-06, + "loss": 0.0126, + "step": 80250 + }, + { + "epoch": 23.35176025603724, + "grad_norm": 0.09742997586727142, + "learning_rate": 8.131646280569605e-06, + "loss": 0.0117, + "step": 80260 + }, + { + "epoch": 23.354669770148384, + "grad_norm": 0.08142542093992233, + "learning_rate": 8.1279716686652e-06, + "loss": 0.0117, + "step": 80270 + }, + { + "epoch": 23.35757928425953, + "grad_norm": 0.09734860807657242, + "learning_rate": 8.124297578645989e-06, + "loss": 0.0115, + "step": 80280 + }, + { + "epoch": 23.36048879837067, + "grad_norm": 0.10368160158395767, + "learning_rate": 8.120624010791002e-06, + "loss": 0.0142, + "step": 80290 + }, + { + "epoch": 23.363398312481817, + "grad_norm": 0.1619454324245453, + "learning_rate": 8.116950965379225e-06, + "loss": 0.0101, + "step": 80300 + }, + { + "epoch": 23.36630782659296, + "grad_norm": 0.09248818457126617, + "learning_rate": 8.113278442689593e-06, + "loss": 0.0112, + "step": 80310 + }, + { + "epoch": 23.369217340704104, + "grad_norm": 0.11162950843572617, + "learning_rate": 8.109606443001019e-06, + "loss": 0.0126, + "step": 80320 + }, + { + "epoch": 23.372126854815246, + "grad_norm": 0.1364763081073761, + "learning_rate": 8.10593496659236e-06, + "loss": 0.014, + "step": 80330 + }, + { + "epoch": 23.375036368926388, + "grad_norm": 0.113436259329319, + "learning_rate": 8.102264013742442e-06, + "loss": 0.0125, + "step": 80340 + }, + { + "epoch": 23.377945883037533, + "grad_norm": 0.11578390002250671, + "learning_rate": 8.098593584730057e-06, + "loss": 0.0118, + "step": 80350 + }, + { + "epoch": 23.380855397148675, + "grad_norm": 0.11272793263196945, + "learning_rate": 8.09492367983394e-06, + "loss": 0.0135, + "step": 80360 + }, + { + "epoch": 23.38376491125982, + "grad_norm": 0.09545538574457169, + "learning_rate": 8.091254299332803e-06, + "loss": 0.0146, + "step": 80370 + }, + { + "epoch": 23.386674425370963, + "grad_norm": 0.10459403693675995, + "learning_rate": 8.087585443505305e-06, + "loss": 0.013, + "step": 80380 + }, + { + "epoch": 23.389583939482108, + "grad_norm": 0.1124049723148346, + "learning_rate": 8.083917112630078e-06, + "loss": 0.0129, + "step": 80390 + }, + { + "epoch": 23.39249345359325, + "grad_norm": 0.34668007493019104, + "learning_rate": 8.0802493069857e-06, + "loss": 0.012, + "step": 80400 + }, + { + "epoch": 23.395402967704392, + "grad_norm": 0.08212853223085403, + "learning_rate": 8.07658202685072e-06, + "loss": 0.0152, + "step": 80410 + }, + { + "epoch": 23.398312481815537, + "grad_norm": 0.10383888334035873, + "learning_rate": 8.072915272503651e-06, + "loss": 0.0117, + "step": 80420 + }, + { + "epoch": 23.40122199592668, + "grad_norm": 0.12535853683948517, + "learning_rate": 8.069249044222951e-06, + "loss": 0.0152, + "step": 80430 + }, + { + "epoch": 23.404131510037825, + "grad_norm": 0.1187693327665329, + "learning_rate": 8.065583342287044e-06, + "loss": 0.0137, + "step": 80440 + }, + { + "epoch": 23.407041024148967, + "grad_norm": 0.1197381466627121, + "learning_rate": 8.061918166974316e-06, + "loss": 0.0124, + "step": 80450 + }, + { + "epoch": 23.409950538260112, + "grad_norm": 0.09074021875858307, + "learning_rate": 8.058253518563122e-06, + "loss": 0.0145, + "step": 80460 + }, + { + "epoch": 23.412860052371254, + "grad_norm": 0.10440906882286072, + "learning_rate": 8.054589397331757e-06, + "loss": 0.0127, + "step": 80470 + }, + { + "epoch": 23.415769566482396, + "grad_norm": 0.09512275457382202, + "learning_rate": 8.050925803558491e-06, + "loss": 0.0142, + "step": 80480 + }, + { + "epoch": 23.41867908059354, + "grad_norm": 0.06672519445419312, + "learning_rate": 8.047262737521556e-06, + "loss": 0.0119, + "step": 80490 + }, + { + "epoch": 23.421588594704684, + "grad_norm": 0.0820726528763771, + "learning_rate": 8.043600199499131e-06, + "loss": 0.0118, + "step": 80500 + }, + { + "epoch": 23.42449810881583, + "grad_norm": 0.10173840820789337, + "learning_rate": 8.03993818976936e-06, + "loss": 0.0133, + "step": 80510 + }, + { + "epoch": 23.42740762292697, + "grad_norm": 0.06551963835954666, + "learning_rate": 8.03627670861035e-06, + "loss": 0.0111, + "step": 80520 + }, + { + "epoch": 23.430317137038113, + "grad_norm": 0.08851084858179092, + "learning_rate": 8.03261575630017e-06, + "loss": 0.011, + "step": 80530 + }, + { + "epoch": 23.43322665114926, + "grad_norm": 0.11487096548080444, + "learning_rate": 8.028955333116846e-06, + "loss": 0.011, + "step": 80540 + }, + { + "epoch": 23.4361361652604, + "grad_norm": 0.08430162072181702, + "learning_rate": 8.025295439338347e-06, + "loss": 0.0119, + "step": 80550 + }, + { + "epoch": 23.439045679371546, + "grad_norm": 0.06152086704969406, + "learning_rate": 8.021636075242641e-06, + "loss": 0.0122, + "step": 80560 + }, + { + "epoch": 23.441955193482688, + "grad_norm": 0.11052606254816055, + "learning_rate": 8.017977241107622e-06, + "loss": 0.0136, + "step": 80570 + }, + { + "epoch": 23.444864707593833, + "grad_norm": 0.1052686870098114, + "learning_rate": 8.01431893721115e-06, + "loss": 0.012, + "step": 80580 + }, + { + "epoch": 23.447774221704975, + "grad_norm": 0.10118360817432404, + "learning_rate": 8.01066116383105e-06, + "loss": 0.014, + "step": 80590 + }, + { + "epoch": 23.450683735816117, + "grad_norm": 0.10555325448513031, + "learning_rate": 8.00700392124512e-06, + "loss": 0.013, + "step": 80600 + }, + { + "epoch": 23.453593249927263, + "grad_norm": 0.06563732028007507, + "learning_rate": 8.003347209731084e-06, + "loss": 0.0122, + "step": 80610 + }, + { + "epoch": 23.456502764038405, + "grad_norm": 0.1160956397652626, + "learning_rate": 7.999691029566658e-06, + "loss": 0.0125, + "step": 80620 + }, + { + "epoch": 23.45941227814955, + "grad_norm": 0.08643034845590591, + "learning_rate": 7.996035381029494e-06, + "loss": 0.0135, + "step": 80630 + }, + { + "epoch": 23.462321792260692, + "grad_norm": 0.09050534665584564, + "learning_rate": 7.99238026439723e-06, + "loss": 0.0124, + "step": 80640 + }, + { + "epoch": 23.465231306371837, + "grad_norm": 0.07935993373394012, + "learning_rate": 7.988725679947432e-06, + "loss": 0.0129, + "step": 80650 + }, + { + "epoch": 23.46814082048298, + "grad_norm": 0.11728908121585846, + "learning_rate": 7.985071627957645e-06, + "loss": 0.0122, + "step": 80660 + }, + { + "epoch": 23.47105033459412, + "grad_norm": 0.18306072056293488, + "learning_rate": 7.981418108705381e-06, + "loss": 0.0131, + "step": 80670 + }, + { + "epoch": 23.473959848705267, + "grad_norm": 0.10635816305875778, + "learning_rate": 7.977765122468089e-06, + "loss": 0.0119, + "step": 80680 + }, + { + "epoch": 23.47686936281641, + "grad_norm": 0.13399194180965424, + "learning_rate": 7.974112669523197e-06, + "loss": 0.0164, + "step": 80690 + }, + { + "epoch": 23.479778876927554, + "grad_norm": 0.12431732565164566, + "learning_rate": 7.970460750148077e-06, + "loss": 0.0112, + "step": 80700 + }, + { + "epoch": 23.482688391038696, + "grad_norm": 0.12431129813194275, + "learning_rate": 7.96680936462007e-06, + "loss": 0.0143, + "step": 80710 + }, + { + "epoch": 23.48559790514984, + "grad_norm": 0.09648348391056061, + "learning_rate": 7.963158513216482e-06, + "loss": 0.0105, + "step": 80720 + }, + { + "epoch": 23.488507419260984, + "grad_norm": 0.11739730834960938, + "learning_rate": 7.959508196214568e-06, + "loss": 0.0131, + "step": 80730 + }, + { + "epoch": 23.491416933372125, + "grad_norm": 0.109571672976017, + "learning_rate": 7.955858413891536e-06, + "loss": 0.0124, + "step": 80740 + }, + { + "epoch": 23.49432644748327, + "grad_norm": 0.09891501069068909, + "learning_rate": 7.952209166524571e-06, + "loss": 0.0123, + "step": 80750 + }, + { + "epoch": 23.497235961594413, + "grad_norm": 0.11239849030971527, + "learning_rate": 7.948560454390814e-06, + "loss": 0.0114, + "step": 80760 + }, + { + "epoch": 23.50014547570556, + "grad_norm": 0.08035823702812195, + "learning_rate": 7.94491227776735e-06, + "loss": 0.0106, + "step": 80770 + }, + { + "epoch": 23.5030549898167, + "grad_norm": 0.1137743890285492, + "learning_rate": 7.941264636931237e-06, + "loss": 0.0125, + "step": 80780 + }, + { + "epoch": 23.505964503927846, + "grad_norm": 0.11350683122873306, + "learning_rate": 7.9376175321595e-06, + "loss": 0.0119, + "step": 80790 + }, + { + "epoch": 23.508874018038988, + "grad_norm": 0.15854588150978088, + "learning_rate": 7.933970963729102e-06, + "loss": 0.0111, + "step": 80800 + }, + { + "epoch": 23.51178353215013, + "grad_norm": 0.09225080907344818, + "learning_rate": 7.93032493191697e-06, + "loss": 0.0145, + "step": 80810 + }, + { + "epoch": 23.514693046261275, + "grad_norm": 0.07517179846763611, + "learning_rate": 7.926679437000006e-06, + "loss": 0.0129, + "step": 80820 + }, + { + "epoch": 23.517602560372417, + "grad_norm": 0.09395122528076172, + "learning_rate": 7.923034479255065e-06, + "loss": 0.014, + "step": 80830 + }, + { + "epoch": 23.520512074483563, + "grad_norm": 0.0953364223241806, + "learning_rate": 7.91939005895895e-06, + "loss": 0.0132, + "step": 80840 + }, + { + "epoch": 23.523421588594704, + "grad_norm": 0.1388244777917862, + "learning_rate": 7.91574617638842e-06, + "loss": 0.0142, + "step": 80850 + }, + { + "epoch": 23.526331102705846, + "grad_norm": 0.10685098171234131, + "learning_rate": 7.912102831820231e-06, + "loss": 0.0143, + "step": 80860 + }, + { + "epoch": 23.529240616816992, + "grad_norm": 0.09484872967004776, + "learning_rate": 7.908460025531051e-06, + "loss": 0.0138, + "step": 80870 + }, + { + "epoch": 23.532150130928134, + "grad_norm": 0.0918048843741417, + "learning_rate": 7.90481775779753e-06, + "loss": 0.0136, + "step": 80880 + }, + { + "epoch": 23.53505964503928, + "grad_norm": 0.15061822533607483, + "learning_rate": 7.901176028896274e-06, + "loss": 0.0149, + "step": 80890 + }, + { + "epoch": 23.53796915915042, + "grad_norm": 0.0951741561293602, + "learning_rate": 7.897534839103856e-06, + "loss": 0.0118, + "step": 80900 + }, + { + "epoch": 23.540878673261567, + "grad_norm": 0.08154623210430145, + "learning_rate": 7.893894188696797e-06, + "loss": 0.0144, + "step": 80910 + }, + { + "epoch": 23.54378818737271, + "grad_norm": 0.10864793509244919, + "learning_rate": 7.89025407795157e-06, + "loss": 0.0132, + "step": 80920 + }, + { + "epoch": 23.54669770148385, + "grad_norm": 0.12371568381786346, + "learning_rate": 7.886614507144628e-06, + "loss": 0.0128, + "step": 80930 + }, + { + "epoch": 23.549607215594996, + "grad_norm": 0.13036499917507172, + "learning_rate": 7.882975476552372e-06, + "loss": 0.0135, + "step": 80940 + }, + { + "epoch": 23.552516729706138, + "grad_norm": 0.10305178165435791, + "learning_rate": 7.87933698645116e-06, + "loss": 0.0104, + "step": 80950 + }, + { + "epoch": 23.555426243817283, + "grad_norm": 0.09468064457178116, + "learning_rate": 7.875699037117306e-06, + "loss": 0.0131, + "step": 80960 + }, + { + "epoch": 23.558335757928425, + "grad_norm": 0.12249180674552917, + "learning_rate": 7.872061628827103e-06, + "loss": 0.0132, + "step": 80970 + }, + { + "epoch": 23.56124527203957, + "grad_norm": 0.0880570113658905, + "learning_rate": 7.86842476185677e-06, + "loss": 0.0103, + "step": 80980 + }, + { + "epoch": 23.564154786150713, + "grad_norm": 0.12612831592559814, + "learning_rate": 7.864788436482522e-06, + "loss": 0.0123, + "step": 80990 + }, + { + "epoch": 23.567064300261855, + "grad_norm": 0.12270201742649078, + "learning_rate": 7.861152652980497e-06, + "loss": 0.0121, + "step": 81000 + }, + { + "epoch": 23.569973814373, + "grad_norm": 0.10333582013845444, + "learning_rate": 7.85751741162682e-06, + "loss": 0.0098, + "step": 81010 + }, + { + "epoch": 23.572883328484142, + "grad_norm": 0.15205809473991394, + "learning_rate": 7.853882712697554e-06, + "loss": 0.0144, + "step": 81020 + }, + { + "epoch": 23.575792842595288, + "grad_norm": 0.09015808999538422, + "learning_rate": 7.850248556468741e-06, + "loss": 0.0109, + "step": 81030 + }, + { + "epoch": 23.57870235670643, + "grad_norm": 0.11366626620292664, + "learning_rate": 7.846614943216361e-06, + "loss": 0.0121, + "step": 81040 + }, + { + "epoch": 23.581611870817575, + "grad_norm": 0.08618906885385513, + "learning_rate": 7.842981873216368e-06, + "loss": 0.0137, + "step": 81050 + }, + { + "epoch": 23.584521384928717, + "grad_norm": 0.11496937274932861, + "learning_rate": 7.839349346744675e-06, + "loss": 0.0144, + "step": 81060 + }, + { + "epoch": 23.58743089903986, + "grad_norm": 0.11915834993124008, + "learning_rate": 7.835717364077138e-06, + "loss": 0.014, + "step": 81070 + }, + { + "epoch": 23.590340413151004, + "grad_norm": 0.12172401696443558, + "learning_rate": 7.832085925489584e-06, + "loss": 0.0125, + "step": 81080 + }, + { + "epoch": 23.593249927262146, + "grad_norm": 0.10628852248191833, + "learning_rate": 7.82845503125781e-06, + "loss": 0.0135, + "step": 81090 + }, + { + "epoch": 23.596159441373292, + "grad_norm": 0.11503509432077408, + "learning_rate": 7.824824681657544e-06, + "loss": 0.0128, + "step": 81100 + }, + { + "epoch": 23.599068955484434, + "grad_norm": 0.11923188716173172, + "learning_rate": 7.821194876964489e-06, + "loss": 0.0132, + "step": 81110 + }, + { + "epoch": 23.60197846959558, + "grad_norm": 0.09143807739019394, + "learning_rate": 7.817565617454305e-06, + "loss": 0.0131, + "step": 81120 + }, + { + "epoch": 23.60488798370672, + "grad_norm": 0.07396473735570908, + "learning_rate": 7.813936903402619e-06, + "loss": 0.0122, + "step": 81130 + }, + { + "epoch": 23.607797497817863, + "grad_norm": 0.11674617975950241, + "learning_rate": 7.810308735085001e-06, + "loss": 0.0127, + "step": 81140 + }, + { + "epoch": 23.61070701192901, + "grad_norm": 0.12394711375236511, + "learning_rate": 7.806681112776976e-06, + "loss": 0.0127, + "step": 81150 + }, + { + "epoch": 23.61361652604015, + "grad_norm": 0.14487721025943756, + "learning_rate": 7.803054036754058e-06, + "loss": 0.0124, + "step": 81160 + }, + { + "epoch": 23.616526040151296, + "grad_norm": 0.1078348383307457, + "learning_rate": 7.799427507291694e-06, + "loss": 0.0143, + "step": 81170 + }, + { + "epoch": 23.619435554262438, + "grad_norm": 0.10270193219184875, + "learning_rate": 7.795801524665284e-06, + "loss": 0.0115, + "step": 81180 + }, + { + "epoch": 23.622345068373583, + "grad_norm": 0.06343881785869598, + "learning_rate": 7.792176089150203e-06, + "loss": 0.0129, + "step": 81190 + }, + { + "epoch": 23.625254582484725, + "grad_norm": 0.08612993359565735, + "learning_rate": 7.788551201021787e-06, + "loss": 0.0116, + "step": 81200 + }, + { + "epoch": 23.628164096595867, + "grad_norm": 0.1368905007839203, + "learning_rate": 7.784926860555315e-06, + "loss": 0.014, + "step": 81210 + }, + { + "epoch": 23.631073610707013, + "grad_norm": 0.11004527658224106, + "learning_rate": 7.781303068026029e-06, + "loss": 0.0118, + "step": 81220 + }, + { + "epoch": 23.633983124818155, + "grad_norm": 0.11310701817274094, + "learning_rate": 7.777679823709134e-06, + "loss": 0.0149, + "step": 81230 + }, + { + "epoch": 23.6368926389293, + "grad_norm": 0.09327834844589233, + "learning_rate": 7.7740571278798e-06, + "loss": 0.0136, + "step": 81240 + }, + { + "epoch": 23.639802153040442, + "grad_norm": 0.15093250572681427, + "learning_rate": 7.770434980813133e-06, + "loss": 0.0133, + "step": 81250 + }, + { + "epoch": 23.642711667151584, + "grad_norm": 0.09899720549583435, + "learning_rate": 7.76681338278422e-06, + "loss": 0.0117, + "step": 81260 + }, + { + "epoch": 23.64562118126273, + "grad_norm": 0.11601489782333374, + "learning_rate": 7.7631923340681e-06, + "loss": 0.0128, + "step": 81270 + }, + { + "epoch": 23.64853069537387, + "grad_norm": 0.09682130813598633, + "learning_rate": 7.759571834939765e-06, + "loss": 0.0121, + "step": 81280 + }, + { + "epoch": 23.651440209485017, + "grad_norm": 0.10942123830318451, + "learning_rate": 7.75595188567416e-06, + "loss": 0.0117, + "step": 81290 + }, + { + "epoch": 23.65434972359616, + "grad_norm": 0.08153918385505676, + "learning_rate": 7.7523324865462e-06, + "loss": 0.0131, + "step": 81300 + }, + { + "epoch": 23.657259237707304, + "grad_norm": 0.08599690347909927, + "learning_rate": 7.748713637830768e-06, + "loss": 0.0115, + "step": 81310 + }, + { + "epoch": 23.660168751818446, + "grad_norm": 0.08310934901237488, + "learning_rate": 7.745095339802672e-06, + "loss": 0.0127, + "step": 81320 + }, + { + "epoch": 23.663078265929588, + "grad_norm": 0.10698913782835007, + "learning_rate": 7.741477592736712e-06, + "loss": 0.0126, + "step": 81330 + }, + { + "epoch": 23.665987780040734, + "grad_norm": 0.14655916392803192, + "learning_rate": 7.737860396907623e-06, + "loss": 0.014, + "step": 81340 + }, + { + "epoch": 23.668897294151876, + "grad_norm": 0.07894794642925262, + "learning_rate": 7.73424375259011e-06, + "loss": 0.0116, + "step": 81350 + }, + { + "epoch": 23.67180680826302, + "grad_norm": 0.09374310821294785, + "learning_rate": 7.730627660058841e-06, + "loss": 0.0101, + "step": 81360 + }, + { + "epoch": 23.674716322374163, + "grad_norm": 0.1619149148464203, + "learning_rate": 7.72701211958842e-06, + "loss": 0.0131, + "step": 81370 + }, + { + "epoch": 23.67762583648531, + "grad_norm": 0.0957368016242981, + "learning_rate": 7.723397131453438e-06, + "loss": 0.0119, + "step": 81380 + }, + { + "epoch": 23.68053535059645, + "grad_norm": 0.10419198870658875, + "learning_rate": 7.719782695928416e-06, + "loss": 0.0126, + "step": 81390 + }, + { + "epoch": 23.683444864707592, + "grad_norm": 0.1281127780675888, + "learning_rate": 7.716168813287858e-06, + "loss": 0.0142, + "step": 81400 + }, + { + "epoch": 23.686354378818738, + "grad_norm": 0.10857229679822922, + "learning_rate": 7.712555483806205e-06, + "loss": 0.0141, + "step": 81410 + }, + { + "epoch": 23.68926389292988, + "grad_norm": 0.10917466878890991, + "learning_rate": 7.70894270775787e-06, + "loss": 0.0127, + "step": 81420 + }, + { + "epoch": 23.692173407041025, + "grad_norm": 0.09018818289041519, + "learning_rate": 7.705330485417224e-06, + "loss": 0.0126, + "step": 81430 + }, + { + "epoch": 23.695082921152167, + "grad_norm": 0.15100347995758057, + "learning_rate": 7.70171881705859e-06, + "loss": 0.0146, + "step": 81440 + }, + { + "epoch": 23.697992435263313, + "grad_norm": 0.08587733656167984, + "learning_rate": 7.698107702956232e-06, + "loss": 0.0129, + "step": 81450 + }, + { + "epoch": 23.700901949374455, + "grad_norm": 0.1320592761039734, + "learning_rate": 7.69449714338442e-06, + "loss": 0.0123, + "step": 81460 + }, + { + "epoch": 23.703811463485597, + "grad_norm": 0.1478150337934494, + "learning_rate": 7.690887138617335e-06, + "loss": 0.0133, + "step": 81470 + }, + { + "epoch": 23.706720977596742, + "grad_norm": 0.11000087857246399, + "learning_rate": 7.687277688929134e-06, + "loss": 0.0113, + "step": 81480 + }, + { + "epoch": 23.709630491707884, + "grad_norm": 0.11071797460317612, + "learning_rate": 7.68366879459393e-06, + "loss": 0.0128, + "step": 81490 + }, + { + "epoch": 23.71254000581903, + "grad_norm": 0.10804122686386108, + "learning_rate": 7.680060455885804e-06, + "loss": 0.0136, + "step": 81500 + }, + { + "epoch": 23.71544951993017, + "grad_norm": 0.11369496583938599, + "learning_rate": 7.676452673078779e-06, + "loss": 0.0118, + "step": 81510 + }, + { + "epoch": 23.718359034041313, + "grad_norm": 0.09902388602495193, + "learning_rate": 7.672845446446835e-06, + "loss": 0.0118, + "step": 81520 + }, + { + "epoch": 23.72126854815246, + "grad_norm": 0.1289895921945572, + "learning_rate": 7.669238776263927e-06, + "loss": 0.0134, + "step": 81530 + }, + { + "epoch": 23.7241780622636, + "grad_norm": 0.13489077985286713, + "learning_rate": 7.665632662803961e-06, + "loss": 0.0115, + "step": 81540 + }, + { + "epoch": 23.727087576374746, + "grad_norm": 0.06097550690174103, + "learning_rate": 7.662027106340785e-06, + "loss": 0.0136, + "step": 81550 + }, + { + "epoch": 23.729997090485888, + "grad_norm": 0.09678100794553757, + "learning_rate": 7.658422107148222e-06, + "loss": 0.0098, + "step": 81560 + }, + { + "epoch": 23.732906604597034, + "grad_norm": 0.09136251360177994, + "learning_rate": 7.654817665500058e-06, + "loss": 0.013, + "step": 81570 + }, + { + "epoch": 23.735816118708176, + "grad_norm": 0.09218855947256088, + "learning_rate": 7.651213781670016e-06, + "loss": 0.0127, + "step": 81580 + }, + { + "epoch": 23.738725632819317, + "grad_norm": 0.07712986320257187, + "learning_rate": 7.647610455931784e-06, + "loss": 0.0123, + "step": 81590 + }, + { + "epoch": 23.741635146930463, + "grad_norm": 0.0797911211848259, + "learning_rate": 7.644007688559018e-06, + "loss": 0.0113, + "step": 81600 + }, + { + "epoch": 23.744544661041605, + "grad_norm": 0.09198153764009476, + "learning_rate": 7.64040547982533e-06, + "loss": 0.0126, + "step": 81610 + }, + { + "epoch": 23.74745417515275, + "grad_norm": 0.11342693865299225, + "learning_rate": 7.636803830004267e-06, + "loss": 0.0126, + "step": 81620 + }, + { + "epoch": 23.750363689263892, + "grad_norm": 0.08946854621171951, + "learning_rate": 7.633202739369368e-06, + "loss": 0.0107, + "step": 81630 + }, + { + "epoch": 23.753273203375038, + "grad_norm": 0.10234414041042328, + "learning_rate": 7.6296022081941e-06, + "loss": 0.0114, + "step": 81640 + }, + { + "epoch": 23.75618271748618, + "grad_norm": 0.1319761723279953, + "learning_rate": 7.6260022367519065e-06, + "loss": 0.0124, + "step": 81650 + }, + { + "epoch": 23.75909223159732, + "grad_norm": 0.11398373544216156, + "learning_rate": 7.622402825316175e-06, + "loss": 0.0125, + "step": 81660 + }, + { + "epoch": 23.762001745708467, + "grad_norm": 0.10755864530801773, + "learning_rate": 7.61880397416026e-06, + "loss": 0.0124, + "step": 81670 + }, + { + "epoch": 23.76491125981961, + "grad_norm": 0.12102973461151123, + "learning_rate": 7.61520568355748e-06, + "loss": 0.0115, + "step": 81680 + }, + { + "epoch": 23.767820773930755, + "grad_norm": 0.14041413366794586, + "learning_rate": 7.611607953781084e-06, + "loss": 0.0131, + "step": 81690 + }, + { + "epoch": 23.770730288041896, + "grad_norm": 0.11478327959775925, + "learning_rate": 7.6080107851043116e-06, + "loss": 0.0127, + "step": 81700 + }, + { + "epoch": 23.773639802153042, + "grad_norm": 0.14033357799053192, + "learning_rate": 7.6044141778003314e-06, + "loss": 0.015, + "step": 81710 + }, + { + "epoch": 23.776549316264184, + "grad_norm": 0.11536005139350891, + "learning_rate": 7.600818132142289e-06, + "loss": 0.0141, + "step": 81720 + }, + { + "epoch": 23.779458830375326, + "grad_norm": 0.10734536498785019, + "learning_rate": 7.597222648403283e-06, + "loss": 0.0117, + "step": 81730 + }, + { + "epoch": 23.78236834448647, + "grad_norm": 0.0891343355178833, + "learning_rate": 7.593627726856365e-06, + "loss": 0.0117, + "step": 81740 + }, + { + "epoch": 23.785277858597613, + "grad_norm": 0.09240922331809998, + "learning_rate": 7.5900333677745335e-06, + "loss": 0.0104, + "step": 81750 + }, + { + "epoch": 23.78818737270876, + "grad_norm": 0.10601887851953506, + "learning_rate": 7.586439571430771e-06, + "loss": 0.0137, + "step": 81760 + }, + { + "epoch": 23.7910968868199, + "grad_norm": 0.12168681621551514, + "learning_rate": 7.5828463380980015e-06, + "loss": 0.014, + "step": 81770 + }, + { + "epoch": 23.794006400931046, + "grad_norm": 0.11172056198120117, + "learning_rate": 7.579253668049098e-06, + "loss": 0.0145, + "step": 81780 + }, + { + "epoch": 23.796915915042188, + "grad_norm": 0.1011413112282753, + "learning_rate": 7.575661561556905e-06, + "loss": 0.0116, + "step": 81790 + }, + { + "epoch": 23.79982542915333, + "grad_norm": 0.11835134774446487, + "learning_rate": 7.572070018894225e-06, + "loss": 0.0137, + "step": 81800 + }, + { + "epoch": 23.802734943264475, + "grad_norm": 0.1536318063735962, + "learning_rate": 7.5684790403338096e-06, + "loss": 0.012, + "step": 81810 + }, + { + "epoch": 23.805644457375617, + "grad_norm": 0.0950169786810875, + "learning_rate": 7.564888626148361e-06, + "loss": 0.0118, + "step": 81820 + }, + { + "epoch": 23.808553971486763, + "grad_norm": 0.11085739731788635, + "learning_rate": 7.561298776610552e-06, + "loss": 0.0124, + "step": 81830 + }, + { + "epoch": 23.811463485597905, + "grad_norm": 0.11517464369535446, + "learning_rate": 7.557709491993017e-06, + "loss": 0.0146, + "step": 81840 + }, + { + "epoch": 23.81437299970905, + "grad_norm": 0.14310887455940247, + "learning_rate": 7.55412077256833e-06, + "loss": 0.0141, + "step": 81850 + }, + { + "epoch": 23.817282513820192, + "grad_norm": 0.0991383045911789, + "learning_rate": 7.550532618609022e-06, + "loss": 0.0119, + "step": 81860 + }, + { + "epoch": 23.820192027931334, + "grad_norm": 0.08636868745088577, + "learning_rate": 7.54694503038761e-06, + "loss": 0.0129, + "step": 81870 + }, + { + "epoch": 23.82310154204248, + "grad_norm": 0.11465732753276825, + "learning_rate": 7.543358008176537e-06, + "loss": 0.0117, + "step": 81880 + }, + { + "epoch": 23.82601105615362, + "grad_norm": 0.15294228494167328, + "learning_rate": 7.539771552248207e-06, + "loss": 0.0104, + "step": 81890 + }, + { + "epoch": 23.828920570264767, + "grad_norm": 0.11894837021827698, + "learning_rate": 7.536185662874997e-06, + "loss": 0.0142, + "step": 81900 + }, + { + "epoch": 23.83183008437591, + "grad_norm": 0.11275901645421982, + "learning_rate": 7.532600340329232e-06, + "loss": 0.0119, + "step": 81910 + }, + { + "epoch": 23.834739598487054, + "grad_norm": 0.08029196411371231, + "learning_rate": 7.529015584883188e-06, + "loss": 0.0116, + "step": 81920 + }, + { + "epoch": 23.837649112598196, + "grad_norm": 0.11938098818063736, + "learning_rate": 7.5254313968091095e-06, + "loss": 0.0127, + "step": 81930 + }, + { + "epoch": 23.84055862670934, + "grad_norm": 0.09227802604436874, + "learning_rate": 7.521847776379183e-06, + "loss": 0.0134, + "step": 81940 + }, + { + "epoch": 23.843468140820484, + "grad_norm": 0.10166507214307785, + "learning_rate": 7.518264723865572e-06, + "loss": 0.0124, + "step": 81950 + }, + { + "epoch": 23.846377654931626, + "grad_norm": 0.13323524594306946, + "learning_rate": 7.514682239540374e-06, + "loss": 0.0135, + "step": 81960 + }, + { + "epoch": 23.84928716904277, + "grad_norm": 0.18477675318717957, + "learning_rate": 7.5111003236756625e-06, + "loss": 0.0141, + "step": 81970 + }, + { + "epoch": 23.852196683153913, + "grad_norm": 0.11462473124265671, + "learning_rate": 7.507518976543463e-06, + "loss": 0.0121, + "step": 81980 + }, + { + "epoch": 23.855106197265055, + "grad_norm": 0.14673922955989838, + "learning_rate": 7.503938198415746e-06, + "loss": 0.0131, + "step": 81990 + }, + { + "epoch": 23.8580157113762, + "grad_norm": 0.11723819375038147, + "learning_rate": 7.5003579895644596e-06, + "loss": 0.015, + "step": 82000 + }, + { + "epoch": 23.860925225487343, + "grad_norm": 0.1134917214512825, + "learning_rate": 7.496778350261485e-06, + "loss": 0.0135, + "step": 82010 + }, + { + "epoch": 23.863834739598488, + "grad_norm": 0.12128591537475586, + "learning_rate": 7.493199280778683e-06, + "loss": 0.013, + "step": 82020 + }, + { + "epoch": 23.86674425370963, + "grad_norm": 0.13865070044994354, + "learning_rate": 7.489620781387849e-06, + "loss": 0.0142, + "step": 82030 + }, + { + "epoch": 23.869653767820775, + "grad_norm": 0.0896529033780098, + "learning_rate": 7.486042852360758e-06, + "loss": 0.0109, + "step": 82040 + }, + { + "epoch": 23.872563281931917, + "grad_norm": 0.09412191063165665, + "learning_rate": 7.482465493969121e-06, + "loss": 0.0138, + "step": 82050 + }, + { + "epoch": 23.87547279604306, + "grad_norm": 0.1172008290886879, + "learning_rate": 7.4788887064846155e-06, + "loss": 0.0127, + "step": 82060 + }, + { + "epoch": 23.878382310154205, + "grad_norm": 0.1011468768119812, + "learning_rate": 7.4753124901788845e-06, + "loss": 0.012, + "step": 82070 + }, + { + "epoch": 23.881291824265347, + "grad_norm": 0.10312607884407043, + "learning_rate": 7.471736845323507e-06, + "loss": 0.0135, + "step": 82080 + }, + { + "epoch": 23.884201338376492, + "grad_norm": 0.08789291232824326, + "learning_rate": 7.4681617721900324e-06, + "loss": 0.0141, + "step": 82090 + }, + { + "epoch": 23.887110852487634, + "grad_norm": 0.08846236020326614, + "learning_rate": 7.4645872710499705e-06, + "loss": 0.0098, + "step": 82100 + }, + { + "epoch": 23.89002036659878, + "grad_norm": 0.1339399814605713, + "learning_rate": 7.4610133421747765e-06, + "loss": 0.0142, + "step": 82110 + }, + { + "epoch": 23.89292988070992, + "grad_norm": 0.13560190796852112, + "learning_rate": 7.457439985835862e-06, + "loss": 0.0133, + "step": 82120 + }, + { + "epoch": 23.895839394821063, + "grad_norm": 0.08713413774967194, + "learning_rate": 7.453867202304602e-06, + "loss": 0.0122, + "step": 82130 + }, + { + "epoch": 23.89874890893221, + "grad_norm": 0.10207775980234146, + "learning_rate": 7.450294991852334e-06, + "loss": 0.014, + "step": 82140 + }, + { + "epoch": 23.90165842304335, + "grad_norm": 0.10218358039855957, + "learning_rate": 7.446723354750337e-06, + "loss": 0.0114, + "step": 82150 + }, + { + "epoch": 23.904567937154496, + "grad_norm": 0.11706096678972244, + "learning_rate": 7.443152291269844e-06, + "loss": 0.0118, + "step": 82160 + }, + { + "epoch": 23.90747745126564, + "grad_norm": 0.10717874765396118, + "learning_rate": 7.439581801682075e-06, + "loss": 0.0139, + "step": 82170 + }, + { + "epoch": 23.910386965376784, + "grad_norm": 0.09272971004247665, + "learning_rate": 7.436011886258174e-06, + "loss": 0.0118, + "step": 82180 + }, + { + "epoch": 23.913296479487926, + "grad_norm": 0.08789035677909851, + "learning_rate": 7.4324425452692445e-06, + "loss": 0.0119, + "step": 82190 + }, + { + "epoch": 23.916205993599068, + "grad_norm": 0.10102546960115433, + "learning_rate": 7.428873778986364e-06, + "loss": 0.0127, + "step": 82200 + }, + { + "epoch": 23.919115507710213, + "grad_norm": 0.14371797442436218, + "learning_rate": 7.42530558768056e-06, + "loss": 0.012, + "step": 82210 + }, + { + "epoch": 23.922025021821355, + "grad_norm": 0.09243465960025787, + "learning_rate": 7.4217379716228075e-06, + "loss": 0.0129, + "step": 82220 + }, + { + "epoch": 23.9249345359325, + "grad_norm": 0.13050049543380737, + "learning_rate": 7.418170931084039e-06, + "loss": 0.0116, + "step": 82230 + }, + { + "epoch": 23.927844050043642, + "grad_norm": 0.13714571297168732, + "learning_rate": 7.414604466335152e-06, + "loss": 0.0147, + "step": 82240 + }, + { + "epoch": 23.930753564154784, + "grad_norm": 0.131483793258667, + "learning_rate": 7.411038577647002e-06, + "loss": 0.012, + "step": 82250 + }, + { + "epoch": 23.93366307826593, + "grad_norm": 0.1075965166091919, + "learning_rate": 7.407473265290385e-06, + "loss": 0.014, + "step": 82260 + }, + { + "epoch": 23.936572592377072, + "grad_norm": 0.08791981637477875, + "learning_rate": 7.403908529536066e-06, + "loss": 0.0106, + "step": 82270 + }, + { + "epoch": 23.939482106488217, + "grad_norm": 0.14243891835212708, + "learning_rate": 7.40034437065477e-06, + "loss": 0.0125, + "step": 82280 + }, + { + "epoch": 23.94239162059936, + "grad_norm": 0.11222469806671143, + "learning_rate": 7.396780788917161e-06, + "loss": 0.0126, + "step": 82290 + }, + { + "epoch": 23.945301134710505, + "grad_norm": 0.1121295914053917, + "learning_rate": 7.393217784593878e-06, + "loss": 0.0131, + "step": 82300 + }, + { + "epoch": 23.948210648821647, + "grad_norm": 0.09923163801431656, + "learning_rate": 7.389655357955499e-06, + "loss": 0.0118, + "step": 82310 + }, + { + "epoch": 23.95112016293279, + "grad_norm": 0.14223536849021912, + "learning_rate": 7.386093509272578e-06, + "loss": 0.0127, + "step": 82320 + }, + { + "epoch": 23.954029677043934, + "grad_norm": 0.11418648064136505, + "learning_rate": 7.3825322388156016e-06, + "loss": 0.0128, + "step": 82330 + }, + { + "epoch": 23.956939191155076, + "grad_norm": 0.13791611790657043, + "learning_rate": 7.3789715468550365e-06, + "loss": 0.0142, + "step": 82340 + }, + { + "epoch": 23.95984870526622, + "grad_norm": 0.10311667621135712, + "learning_rate": 7.375411433661281e-06, + "loss": 0.0128, + "step": 82350 + }, + { + "epoch": 23.962758219377363, + "grad_norm": 0.1287662386894226, + "learning_rate": 7.371851899504712e-06, + "loss": 0.012, + "step": 82360 + }, + { + "epoch": 23.96566773348851, + "grad_norm": 0.09995871037244797, + "learning_rate": 7.368292944655653e-06, + "loss": 0.0133, + "step": 82370 + }, + { + "epoch": 23.96857724759965, + "grad_norm": 0.07809341698884964, + "learning_rate": 7.3647345693843746e-06, + "loss": 0.0122, + "step": 82380 + }, + { + "epoch": 23.971486761710793, + "grad_norm": 0.11243429780006409, + "learning_rate": 7.361176773961118e-06, + "loss": 0.0117, + "step": 82390 + }, + { + "epoch": 23.974396275821938, + "grad_norm": 0.16213670372962952, + "learning_rate": 7.357619558656079e-06, + "loss": 0.012, + "step": 82400 + }, + { + "epoch": 23.97730578993308, + "grad_norm": 0.09824518859386444, + "learning_rate": 7.3540629237394e-06, + "loss": 0.0154, + "step": 82410 + }, + { + "epoch": 23.980215304044226, + "grad_norm": 0.17611005902290344, + "learning_rate": 7.350506869481177e-06, + "loss": 0.0122, + "step": 82420 + }, + { + "epoch": 23.983124818155368, + "grad_norm": 0.13222116231918335, + "learning_rate": 7.3469513961514745e-06, + "loss": 0.0128, + "step": 82430 + }, + { + "epoch": 23.986034332266513, + "grad_norm": 0.07638631016016006, + "learning_rate": 7.343396504020313e-06, + "loss": 0.0112, + "step": 82440 + }, + { + "epoch": 23.988943846377655, + "grad_norm": 0.15621493756771088, + "learning_rate": 7.339842193357659e-06, + "loss": 0.0142, + "step": 82450 + }, + { + "epoch": 23.991853360488797, + "grad_norm": 0.08129988610744476, + "learning_rate": 7.336288464433427e-06, + "loss": 0.0116, + "step": 82460 + }, + { + "epoch": 23.994762874599942, + "grad_norm": 0.09794914722442627, + "learning_rate": 7.332735317517519e-06, + "loss": 0.0122, + "step": 82470 + }, + { + "epoch": 23.997672388711084, + "grad_norm": 0.10776340961456299, + "learning_rate": 7.329182752879765e-06, + "loss": 0.0131, + "step": 82480 + }, + { + "epoch": 24.00058190282223, + "grad_norm": 0.14134608209133148, + "learning_rate": 7.3256307707899525e-06, + "loss": 0.0125, + "step": 82490 + }, + { + "epoch": 24.00349141693337, + "grad_norm": 0.11423341929912567, + "learning_rate": 7.322079371517838e-06, + "loss": 0.0133, + "step": 82500 + }, + { + "epoch": 24.006400931044517, + "grad_norm": 0.09195499122142792, + "learning_rate": 7.31852855533313e-06, + "loss": 0.0123, + "step": 82510 + }, + { + "epoch": 24.00931044515566, + "grad_norm": 0.12623442709445953, + "learning_rate": 7.314978322505485e-06, + "loss": 0.0122, + "step": 82520 + }, + { + "epoch": 24.0122199592668, + "grad_norm": 0.09605465084314346, + "learning_rate": 7.311428673304513e-06, + "loss": 0.0122, + "step": 82530 + }, + { + "epoch": 24.015129473377947, + "grad_norm": 0.1834714114665985, + "learning_rate": 7.307879607999794e-06, + "loss": 0.0136, + "step": 82540 + }, + { + "epoch": 24.01803898748909, + "grad_norm": 0.10073646157979965, + "learning_rate": 7.304331126860859e-06, + "loss": 0.0131, + "step": 82550 + }, + { + "epoch": 24.020948501600234, + "grad_norm": 0.09505777060985565, + "learning_rate": 7.300783230157185e-06, + "loss": 0.0119, + "step": 82560 + }, + { + "epoch": 24.023858015711376, + "grad_norm": 0.11527670174837112, + "learning_rate": 7.297235918158213e-06, + "loss": 0.0128, + "step": 82570 + }, + { + "epoch": 24.02676752982252, + "grad_norm": 0.10394103080034256, + "learning_rate": 7.293689191133345e-06, + "loss": 0.0122, + "step": 82580 + }, + { + "epoch": 24.029677043933663, + "grad_norm": 0.11201265454292297, + "learning_rate": 7.290143049351926e-06, + "loss": 0.0135, + "step": 82590 + }, + { + "epoch": 24.032586558044805, + "grad_norm": 0.1349404901266098, + "learning_rate": 7.286597493083255e-06, + "loss": 0.0148, + "step": 82600 + }, + { + "epoch": 24.03549607215595, + "grad_norm": 0.10768667608499527, + "learning_rate": 7.283052522596601e-06, + "loss": 0.0123, + "step": 82610 + }, + { + "epoch": 24.038405586267093, + "grad_norm": 0.09394870698451996, + "learning_rate": 7.279508138161188e-06, + "loss": 0.0135, + "step": 82620 + }, + { + "epoch": 24.041315100378238, + "grad_norm": 0.1280023753643036, + "learning_rate": 7.275964340046174e-06, + "loss": 0.014, + "step": 82630 + }, + { + "epoch": 24.04422461448938, + "grad_norm": 0.12544430792331696, + "learning_rate": 7.272421128520698e-06, + "loss": 0.0128, + "step": 82640 + }, + { + "epoch": 24.047134128600522, + "grad_norm": 0.11309689283370972, + "learning_rate": 7.268878503853835e-06, + "loss": 0.0125, + "step": 82650 + }, + { + "epoch": 24.050043642711668, + "grad_norm": 0.09666505455970764, + "learning_rate": 7.265336466314631e-06, + "loss": 0.0119, + "step": 82660 + }, + { + "epoch": 24.05295315682281, + "grad_norm": 0.09268681704998016, + "learning_rate": 7.261795016172082e-06, + "loss": 0.0125, + "step": 82670 + }, + { + "epoch": 24.055862670933955, + "grad_norm": 0.10470244288444519, + "learning_rate": 7.258254153695128e-06, + "loss": 0.0122, + "step": 82680 + }, + { + "epoch": 24.058772185045097, + "grad_norm": 0.10435837507247925, + "learning_rate": 7.2547138791526855e-06, + "loss": 0.0173, + "step": 82690 + }, + { + "epoch": 24.061681699156242, + "grad_norm": 0.14002597332000732, + "learning_rate": 7.2511741928136045e-06, + "loss": 0.014, + "step": 82700 + }, + { + "epoch": 24.064591213267384, + "grad_norm": 0.10739494860172272, + "learning_rate": 7.24763509494671e-06, + "loss": 0.013, + "step": 82710 + }, + { + "epoch": 24.067500727378526, + "grad_norm": 0.1249949187040329, + "learning_rate": 7.244096585820764e-06, + "loss": 0.0108, + "step": 82720 + }, + { + "epoch": 24.07041024148967, + "grad_norm": 0.13841913640499115, + "learning_rate": 7.240558665704498e-06, + "loss": 0.0115, + "step": 82730 + }, + { + "epoch": 24.073319755600814, + "grad_norm": 0.11475188285112381, + "learning_rate": 7.237021334866599e-06, + "loss": 0.0119, + "step": 82740 + }, + { + "epoch": 24.07622926971196, + "grad_norm": 0.12822413444519043, + "learning_rate": 7.2334845935757e-06, + "loss": 0.014, + "step": 82750 + }, + { + "epoch": 24.0791387838231, + "grad_norm": 0.11842960119247437, + "learning_rate": 7.229948442100379e-06, + "loss": 0.0129, + "step": 82760 + }, + { + "epoch": 24.082048297934247, + "grad_norm": 0.09881658852100372, + "learning_rate": 7.2264128807092055e-06, + "loss": 0.0115, + "step": 82770 + }, + { + "epoch": 24.08495781204539, + "grad_norm": 0.21406611800193787, + "learning_rate": 7.222877909670674e-06, + "loss": 0.0106, + "step": 82780 + }, + { + "epoch": 24.08786732615653, + "grad_norm": 0.10350848734378815, + "learning_rate": 7.219343529253235e-06, + "loss": 0.0127, + "step": 82790 + }, + { + "epoch": 24.090776840267676, + "grad_norm": 0.11177726835012436, + "learning_rate": 7.215809739725308e-06, + "loss": 0.0132, + "step": 82800 + }, + { + "epoch": 24.093686354378818, + "grad_norm": 0.1693701297044754, + "learning_rate": 7.212276541355266e-06, + "loss": 0.0161, + "step": 82810 + }, + { + "epoch": 24.096595868489963, + "grad_norm": 0.09501361846923828, + "learning_rate": 7.208743934411424e-06, + "loss": 0.0133, + "step": 82820 + }, + { + "epoch": 24.099505382601105, + "grad_norm": 0.11690454185009003, + "learning_rate": 7.205211919162058e-06, + "loss": 0.0115, + "step": 82830 + }, + { + "epoch": 24.10241489671225, + "grad_norm": 0.14387734234333038, + "learning_rate": 7.201680495875405e-06, + "loss": 0.0141, + "step": 82840 + }, + { + "epoch": 24.105324410823393, + "grad_norm": 0.12916497886180878, + "learning_rate": 7.1981496648196606e-06, + "loss": 0.0116, + "step": 82850 + }, + { + "epoch": 24.108233924934535, + "grad_norm": 0.165337935090065, + "learning_rate": 7.194619426262961e-06, + "loss": 0.0116, + "step": 82860 + }, + { + "epoch": 24.11114343904568, + "grad_norm": 0.10101848840713501, + "learning_rate": 7.191089780473393e-06, + "loss": 0.0122, + "step": 82870 + }, + { + "epoch": 24.114052953156822, + "grad_norm": 0.07546888291835785, + "learning_rate": 7.187560727719033e-06, + "loss": 0.0124, + "step": 82880 + }, + { + "epoch": 24.116962467267967, + "grad_norm": 0.1076006069779396, + "learning_rate": 7.184032268267879e-06, + "loss": 0.0147, + "step": 82890 + }, + { + "epoch": 24.11987198137911, + "grad_norm": 0.09824365377426147, + "learning_rate": 7.180504402387888e-06, + "loss": 0.0133, + "step": 82900 + }, + { + "epoch": 24.122781495490255, + "grad_norm": 0.10781807452440262, + "learning_rate": 7.176977130346983e-06, + "loss": 0.0135, + "step": 82910 + }, + { + "epoch": 24.125691009601397, + "grad_norm": 0.11603430658578873, + "learning_rate": 7.173450452413043e-06, + "loss": 0.0119, + "step": 82920 + }, + { + "epoch": 24.12860052371254, + "grad_norm": 0.09247943758964539, + "learning_rate": 7.169924368853884e-06, + "loss": 0.0122, + "step": 82930 + }, + { + "epoch": 24.131510037823684, + "grad_norm": 0.1023123636841774, + "learning_rate": 7.166398879937302e-06, + "loss": 0.012, + "step": 82940 + }, + { + "epoch": 24.134419551934826, + "grad_norm": 0.09903573989868164, + "learning_rate": 7.162873985931023e-06, + "loss": 0.0119, + "step": 82950 + }, + { + "epoch": 24.13732906604597, + "grad_norm": 0.15753623843193054, + "learning_rate": 7.159349687102748e-06, + "loss": 0.0124, + "step": 82960 + }, + { + "epoch": 24.140238580157114, + "grad_norm": 0.09723465144634247, + "learning_rate": 7.155825983720115e-06, + "loss": 0.0117, + "step": 82970 + }, + { + "epoch": 24.143148094268255, + "grad_norm": 0.09602390974760056, + "learning_rate": 7.152302876050733e-06, + "loss": 0.012, + "step": 82980 + }, + { + "epoch": 24.1460576083794, + "grad_norm": 0.12413492798805237, + "learning_rate": 7.148780364362162e-06, + "loss": 0.0141, + "step": 82990 + }, + { + "epoch": 24.148967122490543, + "grad_norm": 0.10741341859102249, + "learning_rate": 7.145258448921903e-06, + "loss": 0.0135, + "step": 83000 + }, + { + "epoch": 24.15187663660169, + "grad_norm": 0.11812126636505127, + "learning_rate": 7.141737129997433e-06, + "loss": 0.0119, + "step": 83010 + }, + { + "epoch": 24.15478615071283, + "grad_norm": 0.15825524926185608, + "learning_rate": 7.1382164078561636e-06, + "loss": 0.0122, + "step": 83020 + }, + { + "epoch": 24.157695664823976, + "grad_norm": 0.074445441365242, + "learning_rate": 7.134696282765473e-06, + "loss": 0.0109, + "step": 83030 + }, + { + "epoch": 24.160605178935118, + "grad_norm": 0.1313764899969101, + "learning_rate": 7.131176754992701e-06, + "loss": 0.0102, + "step": 83040 + }, + { + "epoch": 24.16351469304626, + "grad_norm": 0.12811067700386047, + "learning_rate": 7.1276578248051225e-06, + "loss": 0.0131, + "step": 83050 + }, + { + "epoch": 24.166424207157405, + "grad_norm": 0.09585830569267273, + "learning_rate": 7.124139492469977e-06, + "loss": 0.0115, + "step": 83060 + }, + { + "epoch": 24.169333721268547, + "grad_norm": 0.09484895318746567, + "learning_rate": 7.120621758254458e-06, + "loss": 0.0115, + "step": 83070 + }, + { + "epoch": 24.172243235379693, + "grad_norm": 0.10908060520887375, + "learning_rate": 7.117104622425724e-06, + "loss": 0.012, + "step": 83080 + }, + { + "epoch": 24.175152749490834, + "grad_norm": 0.08403359353542328, + "learning_rate": 7.113588085250865e-06, + "loss": 0.0134, + "step": 83090 + }, + { + "epoch": 24.17806226360198, + "grad_norm": 0.1226910948753357, + "learning_rate": 7.1100721469969445e-06, + "loss": 0.0111, + "step": 83100 + }, + { + "epoch": 24.180971777713122, + "grad_norm": 0.11065986752510071, + "learning_rate": 7.106556807930983e-06, + "loss": 0.0155, + "step": 83110 + }, + { + "epoch": 24.183881291824264, + "grad_norm": 0.08238428086042404, + "learning_rate": 7.103042068319938e-06, + "loss": 0.0112, + "step": 83120 + }, + { + "epoch": 24.18679080593541, + "grad_norm": 0.1462012678384781, + "learning_rate": 7.099527928430727e-06, + "loss": 0.0137, + "step": 83130 + }, + { + "epoch": 24.18970032004655, + "grad_norm": 0.09594299644231796, + "learning_rate": 7.096014388530232e-06, + "loss": 0.0108, + "step": 83140 + }, + { + "epoch": 24.192609834157697, + "grad_norm": 0.10113846510648727, + "learning_rate": 7.092501448885289e-06, + "loss": 0.0116, + "step": 83150 + }, + { + "epoch": 24.19551934826884, + "grad_norm": 0.09511218965053558, + "learning_rate": 7.088989109762675e-06, + "loss": 0.0118, + "step": 83160 + }, + { + "epoch": 24.198428862379984, + "grad_norm": 0.11964470148086548, + "learning_rate": 7.085477371429119e-06, + "loss": 0.0133, + "step": 83170 + }, + { + "epoch": 24.201338376491126, + "grad_norm": 0.1072021871805191, + "learning_rate": 7.081966234151336e-06, + "loss": 0.0133, + "step": 83180 + }, + { + "epoch": 24.204247890602268, + "grad_norm": 0.07359758764505386, + "learning_rate": 7.078455698195963e-06, + "loss": 0.0117, + "step": 83190 + }, + { + "epoch": 24.207157404713413, + "grad_norm": 0.08812130987644196, + "learning_rate": 7.0749457638295975e-06, + "loss": 0.0127, + "step": 83200 + }, + { + "epoch": 24.210066918824555, + "grad_norm": 0.09301616996526718, + "learning_rate": 7.071436431318798e-06, + "loss": 0.011, + "step": 83210 + }, + { + "epoch": 24.2129764329357, + "grad_norm": 0.12970015406608582, + "learning_rate": 7.067927700930085e-06, + "loss": 0.0115, + "step": 83220 + }, + { + "epoch": 24.215885947046843, + "grad_norm": 0.17245851457118988, + "learning_rate": 7.064419572929911e-06, + "loss": 0.0112, + "step": 83230 + }, + { + "epoch": 24.21879546115799, + "grad_norm": 0.07814918458461761, + "learning_rate": 7.060912047584704e-06, + "loss": 0.0125, + "step": 83240 + }, + { + "epoch": 24.22170497526913, + "grad_norm": 0.1193484514951706, + "learning_rate": 7.057405125160829e-06, + "loss": 0.0141, + "step": 83250 + }, + { + "epoch": 24.224614489380272, + "grad_norm": 0.08192596584558487, + "learning_rate": 7.053898805924625e-06, + "loss": 0.0107, + "step": 83260 + }, + { + "epoch": 24.227524003491418, + "grad_norm": 0.14677855372428894, + "learning_rate": 7.0503930901423596e-06, + "loss": 0.0116, + "step": 83270 + }, + { + "epoch": 24.23043351760256, + "grad_norm": 0.1094379872083664, + "learning_rate": 7.046887978080276e-06, + "loss": 0.0116, + "step": 83280 + }, + { + "epoch": 24.233343031713705, + "grad_norm": 0.17159543931484222, + "learning_rate": 7.04338347000457e-06, + "loss": 0.0132, + "step": 83290 + }, + { + "epoch": 24.236252545824847, + "grad_norm": 0.10508960485458374, + "learning_rate": 7.039879566181377e-06, + "loss": 0.0134, + "step": 83300 + }, + { + "epoch": 24.23916205993599, + "grad_norm": 0.09240979701280594, + "learning_rate": 7.036376266876802e-06, + "loss": 0.0141, + "step": 83310 + }, + { + "epoch": 24.242071574047134, + "grad_norm": 0.13323789834976196, + "learning_rate": 7.03287357235689e-06, + "loss": 0.0133, + "step": 83320 + }, + { + "epoch": 24.244981088158276, + "grad_norm": 0.09417569637298584, + "learning_rate": 7.029371482887656e-06, + "loss": 0.0122, + "step": 83330 + }, + { + "epoch": 24.247890602269422, + "grad_norm": 0.11300661414861679, + "learning_rate": 7.025869998735052e-06, + "loss": 0.0118, + "step": 83340 + }, + { + "epoch": 24.250800116380564, + "grad_norm": 0.09162361919879913, + "learning_rate": 7.022369120165003e-06, + "loss": 0.0128, + "step": 83350 + }, + { + "epoch": 24.25370963049171, + "grad_norm": 0.09871980547904968, + "learning_rate": 7.018868847443365e-06, + "loss": 0.0127, + "step": 83360 + }, + { + "epoch": 24.25661914460285, + "grad_norm": 0.15046939253807068, + "learning_rate": 7.0153691808359675e-06, + "loss": 0.0109, + "step": 83370 + }, + { + "epoch": 24.259528658713993, + "grad_norm": 0.10397671908140182, + "learning_rate": 7.01187012060859e-06, + "loss": 0.0132, + "step": 83380 + }, + { + "epoch": 24.26243817282514, + "grad_norm": 0.10557131469249725, + "learning_rate": 7.008371667026958e-06, + "loss": 0.012, + "step": 83390 + }, + { + "epoch": 24.26534768693628, + "grad_norm": 0.10859280824661255, + "learning_rate": 7.004873820356754e-06, + "loss": 0.013, + "step": 83400 + }, + { + "epoch": 24.268257201047426, + "grad_norm": 0.10231134295463562, + "learning_rate": 7.0013765808636295e-06, + "loss": 0.0123, + "step": 83410 + }, + { + "epoch": 24.271166715158568, + "grad_norm": 0.08701319992542267, + "learning_rate": 6.9978799488131635e-06, + "loss": 0.0115, + "step": 83420 + }, + { + "epoch": 24.274076229269713, + "grad_norm": 0.10420706123113632, + "learning_rate": 6.994383924470904e-06, + "loss": 0.0128, + "step": 83430 + }, + { + "epoch": 24.276985743380855, + "grad_norm": 0.08345292508602142, + "learning_rate": 6.990888508102351e-06, + "loss": 0.0106, + "step": 83440 + }, + { + "epoch": 24.279895257491997, + "grad_norm": 0.1129554957151413, + "learning_rate": 6.987393699972966e-06, + "loss": 0.0125, + "step": 83450 + }, + { + "epoch": 24.282804771603143, + "grad_norm": 0.13869084417819977, + "learning_rate": 6.98389950034815e-06, + "loss": 0.0109, + "step": 83460 + }, + { + "epoch": 24.285714285714285, + "grad_norm": 0.10070401430130005, + "learning_rate": 6.980405909493255e-06, + "loss": 0.0119, + "step": 83470 + }, + { + "epoch": 24.28862379982543, + "grad_norm": 0.08557243645191193, + "learning_rate": 6.976912927673619e-06, + "loss": 0.0111, + "step": 83480 + }, + { + "epoch": 24.291533313936572, + "grad_norm": 0.15485282242298126, + "learning_rate": 6.973420555154496e-06, + "loss": 0.0104, + "step": 83490 + }, + { + "epoch": 24.294442828047718, + "grad_norm": 0.21515241265296936, + "learning_rate": 6.969928792201108e-06, + "loss": 0.0129, + "step": 83500 + }, + { + "epoch": 24.29735234215886, + "grad_norm": 0.11799183487892151, + "learning_rate": 6.966437639078632e-06, + "loss": 0.0141, + "step": 83510 + }, + { + "epoch": 24.30026185627, + "grad_norm": 0.07658097892999649, + "learning_rate": 6.9629470960522065e-06, + "loss": 0.0111, + "step": 83520 + }, + { + "epoch": 24.303171370381147, + "grad_norm": 0.08752334117889404, + "learning_rate": 6.95945716338691e-06, + "loss": 0.012, + "step": 83530 + }, + { + "epoch": 24.30608088449229, + "grad_norm": 0.10720223933458328, + "learning_rate": 6.955967841347772e-06, + "loss": 0.0126, + "step": 83540 + }, + { + "epoch": 24.308990398603434, + "grad_norm": 0.12938199937343597, + "learning_rate": 6.9524791301997924e-06, + "loss": 0.011, + "step": 83550 + }, + { + "epoch": 24.311899912714576, + "grad_norm": 0.11460397392511368, + "learning_rate": 6.948991030207918e-06, + "loss": 0.013, + "step": 83560 + }, + { + "epoch": 24.31480942682572, + "grad_norm": 0.09389664232730865, + "learning_rate": 6.945503541637038e-06, + "loss": 0.0116, + "step": 83570 + }, + { + "epoch": 24.317718940936864, + "grad_norm": 0.1028117761015892, + "learning_rate": 6.942016664752009e-06, + "loss": 0.0101, + "step": 83580 + }, + { + "epoch": 24.320628455048006, + "grad_norm": 0.1144455298781395, + "learning_rate": 6.938530399817642e-06, + "loss": 0.0127, + "step": 83590 + }, + { + "epoch": 24.32353796915915, + "grad_norm": 0.09642479568719864, + "learning_rate": 6.935044747098684e-06, + "loss": 0.013, + "step": 83600 + }, + { + "epoch": 24.326447483270293, + "grad_norm": 0.18028739094734192, + "learning_rate": 6.931559706859857e-06, + "loss": 0.0106, + "step": 83610 + }, + { + "epoch": 24.32935699738144, + "grad_norm": 0.10355337709188461, + "learning_rate": 6.928075279365821e-06, + "loss": 0.012, + "step": 83620 + }, + { + "epoch": 24.33226651149258, + "grad_norm": 0.10301706939935684, + "learning_rate": 6.924591464881201e-06, + "loss": 0.0139, + "step": 83630 + }, + { + "epoch": 24.335176025603722, + "grad_norm": 0.0841703861951828, + "learning_rate": 6.921108263670561e-06, + "loss": 0.011, + "step": 83640 + }, + { + "epoch": 24.338085539714868, + "grad_norm": 0.11399044841527939, + "learning_rate": 6.917625675998439e-06, + "loss": 0.0112, + "step": 83650 + }, + { + "epoch": 24.34099505382601, + "grad_norm": 0.09500688314437866, + "learning_rate": 6.914143702129301e-06, + "loss": 0.0119, + "step": 83660 + }, + { + "epoch": 24.343904567937155, + "grad_norm": 0.10691593587398529, + "learning_rate": 6.910662342327588e-06, + "loss": 0.0137, + "step": 83670 + }, + { + "epoch": 24.346814082048297, + "grad_norm": 0.16883257031440735, + "learning_rate": 6.907181596857688e-06, + "loss": 0.0129, + "step": 83680 + }, + { + "epoch": 24.349723596159443, + "grad_norm": 0.14022555947303772, + "learning_rate": 6.903701465983935e-06, + "loss": 0.0126, + "step": 83690 + }, + { + "epoch": 24.352633110270585, + "grad_norm": 0.11728333681821823, + "learning_rate": 6.900221949970629e-06, + "loss": 0.0114, + "step": 83700 + }, + { + "epoch": 24.355542624381727, + "grad_norm": 0.11998418718576431, + "learning_rate": 6.896743049082007e-06, + "loss": 0.0127, + "step": 83710 + }, + { + "epoch": 24.358452138492872, + "grad_norm": 0.11773742735385895, + "learning_rate": 6.893264763582279e-06, + "loss": 0.0136, + "step": 83720 + }, + { + "epoch": 24.361361652604014, + "grad_norm": 0.16111917793750763, + "learning_rate": 6.889787093735587e-06, + "loss": 0.0119, + "step": 83730 + }, + { + "epoch": 24.36427116671516, + "grad_norm": 0.08807458728551865, + "learning_rate": 6.886310039806043e-06, + "loss": 0.0104, + "step": 83740 + }, + { + "epoch": 24.3671806808263, + "grad_norm": 0.0880085676908493, + "learning_rate": 6.88283360205771e-06, + "loss": 0.0114, + "step": 83750 + }, + { + "epoch": 24.370090194937447, + "grad_norm": 0.09356711804866791, + "learning_rate": 6.8793577807545966e-06, + "loss": 0.0116, + "step": 83760 + }, + { + "epoch": 24.37299970904859, + "grad_norm": 0.1082274317741394, + "learning_rate": 6.875882576160657e-06, + "loss": 0.0124, + "step": 83770 + }, + { + "epoch": 24.37590922315973, + "grad_norm": 0.0960974246263504, + "learning_rate": 6.872407988539833e-06, + "loss": 0.0131, + "step": 83780 + }, + { + "epoch": 24.378818737270876, + "grad_norm": 0.11708246171474457, + "learning_rate": 6.8689340181559844e-06, + "loss": 0.013, + "step": 83790 + }, + { + "epoch": 24.381728251382018, + "grad_norm": 0.11301849037408829, + "learning_rate": 6.8654606652729315e-06, + "loss": 0.0116, + "step": 83800 + }, + { + "epoch": 24.384637765493164, + "grad_norm": 0.09922720491886139, + "learning_rate": 6.861987930154457e-06, + "loss": 0.0106, + "step": 83810 + }, + { + "epoch": 24.387547279604306, + "grad_norm": 0.12191013246774673, + "learning_rate": 6.8585158130643e-06, + "loss": 0.011, + "step": 83820 + }, + { + "epoch": 24.39045679371545, + "grad_norm": 0.12471474707126617, + "learning_rate": 6.855044314266138e-06, + "loss": 0.0123, + "step": 83830 + }, + { + "epoch": 24.393366307826593, + "grad_norm": 0.13783583045005798, + "learning_rate": 6.851573434023602e-06, + "loss": 0.0105, + "step": 83840 + }, + { + "epoch": 24.396275821937735, + "grad_norm": 0.12884919345378876, + "learning_rate": 6.848103172600287e-06, + "loss": 0.0127, + "step": 83850 + }, + { + "epoch": 24.39918533604888, + "grad_norm": 0.08345529437065125, + "learning_rate": 6.844633530259747e-06, + "loss": 0.0125, + "step": 83860 + }, + { + "epoch": 24.402094850160022, + "grad_norm": 0.09292532503604889, + "learning_rate": 6.841164507265461e-06, + "loss": 0.0116, + "step": 83870 + }, + { + "epoch": 24.405004364271168, + "grad_norm": 0.0853348895907402, + "learning_rate": 6.8376961038808886e-06, + "loss": 0.0132, + "step": 83880 + }, + { + "epoch": 24.40791387838231, + "grad_norm": 0.09345197677612305, + "learning_rate": 6.834228320369436e-06, + "loss": 0.0132, + "step": 83890 + }, + { + "epoch": 24.410823392493455, + "grad_norm": 0.10964822769165039, + "learning_rate": 6.830761156994454e-06, + "loss": 0.0123, + "step": 83900 + }, + { + "epoch": 24.413732906604597, + "grad_norm": 0.07953955233097076, + "learning_rate": 6.827294614019243e-06, + "loss": 0.0112, + "step": 83910 + }, + { + "epoch": 24.41664242071574, + "grad_norm": 0.09500192105770111, + "learning_rate": 6.82382869170707e-06, + "loss": 0.0109, + "step": 83920 + }, + { + "epoch": 24.419551934826885, + "grad_norm": 0.13088753819465637, + "learning_rate": 6.820363390321158e-06, + "loss": 0.0117, + "step": 83930 + }, + { + "epoch": 24.422461448938027, + "grad_norm": 0.10180121660232544, + "learning_rate": 6.816898710124658e-06, + "loss": 0.0124, + "step": 83940 + }, + { + "epoch": 24.425370963049172, + "grad_norm": 0.1419447809457779, + "learning_rate": 6.8134346513807035e-06, + "loss": 0.0121, + "step": 83950 + }, + { + "epoch": 24.428280477160314, + "grad_norm": 0.13770371675491333, + "learning_rate": 6.8099712143523555e-06, + "loss": 0.0135, + "step": 83960 + }, + { + "epoch": 24.43118999127146, + "grad_norm": 0.07930774986743927, + "learning_rate": 6.806508399302644e-06, + "loss": 0.0113, + "step": 83970 + }, + { + "epoch": 24.4340995053826, + "grad_norm": 0.10922575742006302, + "learning_rate": 6.803046206494553e-06, + "loss": 0.0142, + "step": 83980 + }, + { + "epoch": 24.437009019493743, + "grad_norm": 0.10654281824827194, + "learning_rate": 6.7995846361910015e-06, + "loss": 0.0134, + "step": 83990 + }, + { + "epoch": 24.43991853360489, + "grad_norm": 0.12408339232206345, + "learning_rate": 6.796123688654883e-06, + "loss": 0.0115, + "step": 84000 + }, + { + "epoch": 24.44282804771603, + "grad_norm": 0.08873133361339569, + "learning_rate": 6.792663364149025e-06, + "loss": 0.0109, + "step": 84010 + }, + { + "epoch": 24.445737561827176, + "grad_norm": 0.092460498213768, + "learning_rate": 6.789203662936227e-06, + "loss": 0.0125, + "step": 84020 + }, + { + "epoch": 24.448647075938318, + "grad_norm": 0.13622546195983887, + "learning_rate": 6.785744585279219e-06, + "loss": 0.0137, + "step": 84030 + }, + { + "epoch": 24.45155659004946, + "grad_norm": 0.12668123841285706, + "learning_rate": 6.7822861314406986e-06, + "loss": 0.0118, + "step": 84040 + }, + { + "epoch": 24.454466104160606, + "grad_norm": 0.08967998623847961, + "learning_rate": 6.778828301683322e-06, + "loss": 0.0129, + "step": 84050 + }, + { + "epoch": 24.457375618271747, + "grad_norm": 0.14119884371757507, + "learning_rate": 6.775371096269681e-06, + "loss": 0.0162, + "step": 84060 + }, + { + "epoch": 24.460285132382893, + "grad_norm": 0.10737556964159012, + "learning_rate": 6.771914515462319e-06, + "loss": 0.0106, + "step": 84070 + }, + { + "epoch": 24.463194646494035, + "grad_norm": 0.09337040036916733, + "learning_rate": 6.76845855952375e-06, + "loss": 0.0107, + "step": 84080 + }, + { + "epoch": 24.46610416060518, + "grad_norm": 0.10821323841810226, + "learning_rate": 6.765003228716437e-06, + "loss": 0.012, + "step": 84090 + }, + { + "epoch": 24.469013674716322, + "grad_norm": 0.1224534884095192, + "learning_rate": 6.761548523302776e-06, + "loss": 0.012, + "step": 84100 + }, + { + "epoch": 24.471923188827464, + "grad_norm": 0.12908561527729034, + "learning_rate": 6.758094443545136e-06, + "loss": 0.0109, + "step": 84110 + }, + { + "epoch": 24.47483270293861, + "grad_norm": 0.1298784762620926, + "learning_rate": 6.754640989705836e-06, + "loss": 0.0135, + "step": 84120 + }, + { + "epoch": 24.47774221704975, + "grad_norm": 0.10730090737342834, + "learning_rate": 6.75118816204714e-06, + "loss": 0.0124, + "step": 84130 + }, + { + "epoch": 24.480651731160897, + "grad_norm": 0.06932874768972397, + "learning_rate": 6.747735960831258e-06, + "loss": 0.011, + "step": 84140 + }, + { + "epoch": 24.48356124527204, + "grad_norm": 0.10484591126441956, + "learning_rate": 6.74428438632037e-06, + "loss": 0.0127, + "step": 84150 + }, + { + "epoch": 24.486470759383185, + "grad_norm": 0.0943303033709526, + "learning_rate": 6.740833438776605e-06, + "loss": 0.0119, + "step": 84160 + }, + { + "epoch": 24.489380273494326, + "grad_norm": 0.1952827125787735, + "learning_rate": 6.737383118462036e-06, + "loss": 0.0112, + "step": 84170 + }, + { + "epoch": 24.49228978760547, + "grad_norm": 0.1292068064212799, + "learning_rate": 6.733933425638679e-06, + "loss": 0.013, + "step": 84180 + }, + { + "epoch": 24.495199301716614, + "grad_norm": 0.09891398996114731, + "learning_rate": 6.730484360568537e-06, + "loss": 0.0152, + "step": 84190 + }, + { + "epoch": 24.498108815827756, + "grad_norm": 0.10014667361974716, + "learning_rate": 6.727035923513536e-06, + "loss": 0.0131, + "step": 84200 + }, + { + "epoch": 24.5010183299389, + "grad_norm": 0.09826000779867172, + "learning_rate": 6.723588114735552e-06, + "loss": 0.0118, + "step": 84210 + }, + { + "epoch": 24.503927844050043, + "grad_norm": 0.12038980424404144, + "learning_rate": 6.720140934496432e-06, + "loss": 0.0125, + "step": 84220 + }, + { + "epoch": 24.50683735816119, + "grad_norm": 0.09138856828212738, + "learning_rate": 6.716694383057971e-06, + "loss": 0.0114, + "step": 84230 + }, + { + "epoch": 24.50974687227233, + "grad_norm": 0.18325775861740112, + "learning_rate": 6.713248460681903e-06, + "loss": 0.0132, + "step": 84240 + }, + { + "epoch": 24.512656386383473, + "grad_norm": 0.16167372465133667, + "learning_rate": 6.709803167629929e-06, + "loss": 0.0122, + "step": 84250 + }, + { + "epoch": 24.515565900494618, + "grad_norm": 0.12649112939834595, + "learning_rate": 6.70635850416369e-06, + "loss": 0.0142, + "step": 84260 + }, + { + "epoch": 24.51847541460576, + "grad_norm": 0.11901970952749252, + "learning_rate": 6.702914470544795e-06, + "loss": 0.0111, + "step": 84270 + }, + { + "epoch": 24.521384928716905, + "grad_norm": 0.11200658231973648, + "learning_rate": 6.699471067034784e-06, + "loss": 0.0106, + "step": 84280 + }, + { + "epoch": 24.524294442828047, + "grad_norm": 0.11619196087121964, + "learning_rate": 6.696028293895167e-06, + "loss": 0.0102, + "step": 84290 + }, + { + "epoch": 24.527203956939193, + "grad_norm": 0.08826770633459091, + "learning_rate": 6.6925861513874044e-06, + "loss": 0.0118, + "step": 84300 + }, + { + "epoch": 24.530113471050335, + "grad_norm": 0.12257298082113266, + "learning_rate": 6.689144639772895e-06, + "loss": 0.0128, + "step": 84310 + }, + { + "epoch": 24.533022985161477, + "grad_norm": 0.12080123275518417, + "learning_rate": 6.685703759313009e-06, + "loss": 0.0132, + "step": 84320 + }, + { + "epoch": 24.535932499272622, + "grad_norm": 0.0904434472322464, + "learning_rate": 6.682263510269045e-06, + "loss": 0.0109, + "step": 84330 + }, + { + "epoch": 24.538842013383764, + "grad_norm": 0.10573882609605789, + "learning_rate": 6.6788238929022775e-06, + "loss": 0.0119, + "step": 84340 + }, + { + "epoch": 24.54175152749491, + "grad_norm": 0.11669043451547623, + "learning_rate": 6.675384907473927e-06, + "loss": 0.0102, + "step": 84350 + }, + { + "epoch": 24.54466104160605, + "grad_norm": 0.1005919948220253, + "learning_rate": 6.671946554245153e-06, + "loss": 0.0125, + "step": 84360 + }, + { + "epoch": 24.547570555717193, + "grad_norm": 0.11740623414516449, + "learning_rate": 6.668508833477072e-06, + "loss": 0.013, + "step": 84370 + }, + { + "epoch": 24.55048006982834, + "grad_norm": 0.12094961106777191, + "learning_rate": 6.665071745430762e-06, + "loss": 0.0107, + "step": 84380 + }, + { + "epoch": 24.55338958393948, + "grad_norm": 0.06900829076766968, + "learning_rate": 6.661635290367254e-06, + "loss": 0.0106, + "step": 84390 + }, + { + "epoch": 24.556299098050626, + "grad_norm": 0.11806595325469971, + "learning_rate": 6.6581994685475105e-06, + "loss": 0.0107, + "step": 84400 + }, + { + "epoch": 24.55920861216177, + "grad_norm": 0.09764142334461212, + "learning_rate": 6.6547642802324684e-06, + "loss": 0.0103, + "step": 84410 + }, + { + "epoch": 24.562118126272914, + "grad_norm": 0.0978640541434288, + "learning_rate": 6.651329725683009e-06, + "loss": 0.0104, + "step": 84420 + }, + { + "epoch": 24.565027640384056, + "grad_norm": 0.08588864654302597, + "learning_rate": 6.647895805159962e-06, + "loss": 0.0103, + "step": 84430 + }, + { + "epoch": 24.567937154495198, + "grad_norm": 0.2674052119255066, + "learning_rate": 6.644462518924096e-06, + "loss": 0.0138, + "step": 84440 + }, + { + "epoch": 24.570846668606343, + "grad_norm": 0.10221811383962631, + "learning_rate": 6.641029867236173e-06, + "loss": 0.0132, + "step": 84450 + }, + { + "epoch": 24.573756182717485, + "grad_norm": 0.1357974261045456, + "learning_rate": 6.637597850356867e-06, + "loss": 0.0122, + "step": 84460 + }, + { + "epoch": 24.57666569682863, + "grad_norm": 0.098868228495121, + "learning_rate": 6.634166468546816e-06, + "loss": 0.0108, + "step": 84470 + }, + { + "epoch": 24.579575210939772, + "grad_norm": 0.05891802906990051, + "learning_rate": 6.630735722066602e-06, + "loss": 0.013, + "step": 84480 + }, + { + "epoch": 24.582484725050918, + "grad_norm": 0.21239420771598816, + "learning_rate": 6.627305611176788e-06, + "loss": 0.0114, + "step": 84490 + }, + { + "epoch": 24.58539423916206, + "grad_norm": 0.09122376143932343, + "learning_rate": 6.62387613613786e-06, + "loss": 0.0121, + "step": 84500 + }, + { + "epoch": 24.588303753273202, + "grad_norm": 0.09854885190725327, + "learning_rate": 6.620447297210253e-06, + "loss": 0.0121, + "step": 84510 + }, + { + "epoch": 24.591213267384347, + "grad_norm": 0.17235167324543, + "learning_rate": 6.6170190946543765e-06, + "loss": 0.0118, + "step": 84520 + }, + { + "epoch": 24.59412278149549, + "grad_norm": 0.16576963663101196, + "learning_rate": 6.613591528730581e-06, + "loss": 0.0132, + "step": 84530 + }, + { + "epoch": 24.597032295606635, + "grad_norm": 0.08923134952783585, + "learning_rate": 6.610164599699164e-06, + "loss": 0.0128, + "step": 84540 + }, + { + "epoch": 24.599941809717777, + "grad_norm": 0.08264674246311188, + "learning_rate": 6.606738307820366e-06, + "loss": 0.0104, + "step": 84550 + }, + { + "epoch": 24.602851323828922, + "grad_norm": 0.09061755239963531, + "learning_rate": 6.603312653354414e-06, + "loss": 0.0108, + "step": 84560 + }, + { + "epoch": 24.605760837940064, + "grad_norm": 0.09853769838809967, + "learning_rate": 6.599887636561452e-06, + "loss": 0.011, + "step": 84570 + }, + { + "epoch": 24.608670352051206, + "grad_norm": 0.09948179870843887, + "learning_rate": 6.5964632577015824e-06, + "loss": 0.0125, + "step": 84580 + }, + { + "epoch": 24.61157986616235, + "grad_norm": 0.16983796656131744, + "learning_rate": 6.593039517034871e-06, + "loss": 0.0125, + "step": 84590 + }, + { + "epoch": 24.614489380273493, + "grad_norm": 0.19061626493930817, + "learning_rate": 6.5896164148213314e-06, + "loss": 0.0112, + "step": 84600 + }, + { + "epoch": 24.61739889438464, + "grad_norm": 0.10366039723157883, + "learning_rate": 6.586193951320918e-06, + "loss": 0.0117, + "step": 84610 + }, + { + "epoch": 24.62030840849578, + "grad_norm": 0.15226268768310547, + "learning_rate": 6.582772126793552e-06, + "loss": 0.012, + "step": 84620 + }, + { + "epoch": 24.623217922606926, + "grad_norm": 0.08066537976264954, + "learning_rate": 6.579350941499088e-06, + "loss": 0.0125, + "step": 84630 + }, + { + "epoch": 24.62612743671807, + "grad_norm": 0.10176479816436768, + "learning_rate": 6.575930395697356e-06, + "loss": 0.0125, + "step": 84640 + }, + { + "epoch": 24.62903695082921, + "grad_norm": 0.09698158502578735, + "learning_rate": 6.572510489648108e-06, + "loss": 0.0141, + "step": 84650 + }, + { + "epoch": 24.631946464940356, + "grad_norm": 0.13663044571876526, + "learning_rate": 6.569091223611073e-06, + "loss": 0.0137, + "step": 84660 + }, + { + "epoch": 24.634855979051498, + "grad_norm": 0.10159241408109665, + "learning_rate": 6.565672597845927e-06, + "loss": 0.0126, + "step": 84670 + }, + { + "epoch": 24.637765493162643, + "grad_norm": 0.07903256267309189, + "learning_rate": 6.562254612612279e-06, + "loss": 0.0117, + "step": 84680 + }, + { + "epoch": 24.640675007273785, + "grad_norm": 0.09893683344125748, + "learning_rate": 6.558837268169715e-06, + "loss": 0.0125, + "step": 84690 + }, + { + "epoch": 24.64358452138493, + "grad_norm": 0.11926772445440292, + "learning_rate": 6.555420564777748e-06, + "loss": 0.0138, + "step": 84700 + }, + { + "epoch": 24.646494035496072, + "grad_norm": 0.10656420886516571, + "learning_rate": 6.552004502695859e-06, + "loss": 0.0119, + "step": 84710 + }, + { + "epoch": 24.649403549607214, + "grad_norm": 0.13725194334983826, + "learning_rate": 6.548589082183481e-06, + "loss": 0.0124, + "step": 84720 + }, + { + "epoch": 24.65231306371836, + "grad_norm": 0.09096939116716385, + "learning_rate": 6.545174303499989e-06, + "loss": 0.014, + "step": 84730 + }, + { + "epoch": 24.6552225778295, + "grad_norm": 0.07673488557338715, + "learning_rate": 6.541760166904705e-06, + "loss": 0.0114, + "step": 84740 + }, + { + "epoch": 24.658132091940647, + "grad_norm": 0.08832289278507233, + "learning_rate": 6.538346672656919e-06, + "loss": 0.0116, + "step": 84750 + }, + { + "epoch": 24.66104160605179, + "grad_norm": 0.14008720219135284, + "learning_rate": 6.5349338210158654e-06, + "loss": 0.0122, + "step": 84760 + }, + { + "epoch": 24.66395112016293, + "grad_norm": 0.10672137886285782, + "learning_rate": 6.5315216122407176e-06, + "loss": 0.0112, + "step": 84770 + }, + { + "epoch": 24.666860634274077, + "grad_norm": 0.1258135437965393, + "learning_rate": 6.528110046590617e-06, + "loss": 0.0136, + "step": 84780 + }, + { + "epoch": 24.66977014838522, + "grad_norm": 0.12626062333583832, + "learning_rate": 6.5246991243246535e-06, + "loss": 0.0127, + "step": 84790 + }, + { + "epoch": 24.672679662496364, + "grad_norm": 0.0899471864104271, + "learning_rate": 6.52128884570186e-06, + "loss": 0.0107, + "step": 84800 + }, + { + "epoch": 24.675589176607506, + "grad_norm": 0.08660640567541122, + "learning_rate": 6.517879210981218e-06, + "loss": 0.0129, + "step": 84810 + }, + { + "epoch": 24.67849869071865, + "grad_norm": 0.14390166103839874, + "learning_rate": 6.514470220421674e-06, + "loss": 0.0124, + "step": 84820 + }, + { + "epoch": 24.681408204829793, + "grad_norm": 0.0861000046133995, + "learning_rate": 6.511061874282122e-06, + "loss": 0.0146, + "step": 84830 + }, + { + "epoch": 24.684317718940935, + "grad_norm": 0.08985184133052826, + "learning_rate": 6.507654172821397e-06, + "loss": 0.0119, + "step": 84840 + }, + { + "epoch": 24.68722723305208, + "grad_norm": 0.10922583192586899, + "learning_rate": 6.504247116298284e-06, + "loss": 0.0127, + "step": 84850 + }, + { + "epoch": 24.690136747163223, + "grad_norm": 0.08032093942165375, + "learning_rate": 6.500840704971546e-06, + "loss": 0.0109, + "step": 84860 + }, + { + "epoch": 24.693046261274368, + "grad_norm": 0.1039203554391861, + "learning_rate": 6.4974349390998665e-06, + "loss": 0.0119, + "step": 84870 + }, + { + "epoch": 24.69595577538551, + "grad_norm": 0.1324070692062378, + "learning_rate": 6.494029818941885e-06, + "loss": 0.0115, + "step": 84880 + }, + { + "epoch": 24.698865289496656, + "grad_norm": 0.08818351477384567, + "learning_rate": 6.490625344756206e-06, + "loss": 0.0111, + "step": 84890 + }, + { + "epoch": 24.701774803607798, + "grad_norm": 0.1277056336402893, + "learning_rate": 6.4872215168013804e-06, + "loss": 0.0119, + "step": 84900 + }, + { + "epoch": 24.70468431771894, + "grad_norm": 0.12060955911874771, + "learning_rate": 6.4838183353359e-06, + "loss": 0.0131, + "step": 84910 + }, + { + "epoch": 24.707593831830085, + "grad_norm": 0.09583266824483871, + "learning_rate": 6.480415800618211e-06, + "loss": 0.011, + "step": 84920 + }, + { + "epoch": 24.710503345941227, + "grad_norm": 0.11352371424436569, + "learning_rate": 6.477013912906716e-06, + "loss": 0.0113, + "step": 84930 + }, + { + "epoch": 24.713412860052372, + "grad_norm": 0.1432604044675827, + "learning_rate": 6.473612672459774e-06, + "loss": 0.0118, + "step": 84940 + }, + { + "epoch": 24.716322374163514, + "grad_norm": 0.1313677281141281, + "learning_rate": 6.470212079535674e-06, + "loss": 0.0127, + "step": 84950 + }, + { + "epoch": 24.71923188827466, + "grad_norm": 0.09042922407388687, + "learning_rate": 6.466812134392675e-06, + "loss": 0.011, + "step": 84960 + }, + { + "epoch": 24.7221414023858, + "grad_norm": 0.12830260396003723, + "learning_rate": 6.463412837288985e-06, + "loss": 0.0118, + "step": 84970 + }, + { + "epoch": 24.725050916496944, + "grad_norm": 0.1021689772605896, + "learning_rate": 6.4600141884827505e-06, + "loss": 0.0111, + "step": 84980 + }, + { + "epoch": 24.72796043060809, + "grad_norm": 0.1302712857723236, + "learning_rate": 6.456616188232082e-06, + "loss": 0.0117, + "step": 84990 + }, + { + "epoch": 24.73086994471923, + "grad_norm": 0.078711599111557, + "learning_rate": 6.453218836795029e-06, + "loss": 0.0123, + "step": 85000 + }, + { + "epoch": 24.733779458830377, + "grad_norm": 0.07898242771625519, + "learning_rate": 6.449822134429607e-06, + "loss": 0.012, + "step": 85010 + }, + { + "epoch": 24.73668897294152, + "grad_norm": 0.14702893793582916, + "learning_rate": 6.446426081393762e-06, + "loss": 0.0124, + "step": 85020 + }, + { + "epoch": 24.73959848705266, + "grad_norm": 0.16647924482822418, + "learning_rate": 6.443030677945413e-06, + "loss": 0.0158, + "step": 85030 + }, + { + "epoch": 24.742508001163806, + "grad_norm": 0.11932359635829926, + "learning_rate": 6.439635924342405e-06, + "loss": 0.0114, + "step": 85040 + }, + { + "epoch": 24.745417515274948, + "grad_norm": 0.10614825040102005, + "learning_rate": 6.436241820842558e-06, + "loss": 0.0128, + "step": 85050 + }, + { + "epoch": 24.748327029386093, + "grad_norm": 0.1117333173751831, + "learning_rate": 6.432848367703634e-06, + "loss": 0.012, + "step": 85060 + }, + { + "epoch": 24.751236543497235, + "grad_norm": 0.11020760983228683, + "learning_rate": 6.429455565183332e-06, + "loss": 0.013, + "step": 85070 + }, + { + "epoch": 24.75414605760838, + "grad_norm": 0.2211574912071228, + "learning_rate": 6.426063413539318e-06, + "loss": 0.0114, + "step": 85080 + }, + { + "epoch": 24.757055571719523, + "grad_norm": 0.1182866096496582, + "learning_rate": 6.422671913029213e-06, + "loss": 0.011, + "step": 85090 + }, + { + "epoch": 24.759965085830665, + "grad_norm": 0.1298506259918213, + "learning_rate": 6.419281063910568e-06, + "loss": 0.0112, + "step": 85100 + }, + { + "epoch": 24.76287459994181, + "grad_norm": 0.10450176894664764, + "learning_rate": 6.415890866440893e-06, + "loss": 0.0124, + "step": 85110 + }, + { + "epoch": 24.765784114052952, + "grad_norm": 0.13375215232372284, + "learning_rate": 6.412501320877659e-06, + "loss": 0.0106, + "step": 85120 + }, + { + "epoch": 24.768693628164097, + "grad_norm": 0.12171649187803268, + "learning_rate": 6.409112427478281e-06, + "loss": 0.013, + "step": 85130 + }, + { + "epoch": 24.77160314227524, + "grad_norm": 0.12570516765117645, + "learning_rate": 6.4057241865001185e-06, + "loss": 0.0124, + "step": 85140 + }, + { + "epoch": 24.774512656386385, + "grad_norm": 0.08737924695014954, + "learning_rate": 6.402336598200476e-06, + "loss": 0.0113, + "step": 85150 + }, + { + "epoch": 24.777422170497527, + "grad_norm": 0.12750956416130066, + "learning_rate": 6.398949662836642e-06, + "loss": 0.0122, + "step": 85160 + }, + { + "epoch": 24.78033168460867, + "grad_norm": 0.09835779666900635, + "learning_rate": 6.395563380665818e-06, + "loss": 0.01, + "step": 85170 + }, + { + "epoch": 24.783241198719814, + "grad_norm": 0.150889590382576, + "learning_rate": 6.392177751945165e-06, + "loss": 0.0111, + "step": 85180 + }, + { + "epoch": 24.786150712830956, + "grad_norm": 0.08370453119277954, + "learning_rate": 6.388792776931806e-06, + "loss": 0.0129, + "step": 85190 + }, + { + "epoch": 24.7890602269421, + "grad_norm": 0.09563086181879044, + "learning_rate": 6.385408455882811e-06, + "loss": 0.0122, + "step": 85200 + }, + { + "epoch": 24.791969741053244, + "grad_norm": 0.12479813396930695, + "learning_rate": 6.382024789055194e-06, + "loss": 0.0119, + "step": 85210 + }, + { + "epoch": 24.79487925516439, + "grad_norm": 0.10720158368349075, + "learning_rate": 6.378641776705912e-06, + "loss": 0.012, + "step": 85220 + }, + { + "epoch": 24.79778876927553, + "grad_norm": 0.07506361603736877, + "learning_rate": 6.375259419091894e-06, + "loss": 0.0107, + "step": 85230 + }, + { + "epoch": 24.800698283386673, + "grad_norm": 0.0850244089961052, + "learning_rate": 6.371877716470009e-06, + "loss": 0.0114, + "step": 85240 + }, + { + "epoch": 24.80360779749782, + "grad_norm": 0.13011720776557922, + "learning_rate": 6.368496669097066e-06, + "loss": 0.0131, + "step": 85250 + }, + { + "epoch": 24.80651731160896, + "grad_norm": 0.14695890247821808, + "learning_rate": 6.365116277229836e-06, + "loss": 0.0127, + "step": 85260 + }, + { + "epoch": 24.809426825720106, + "grad_norm": 0.10990049690008163, + "learning_rate": 6.361736541125047e-06, + "loss": 0.0136, + "step": 85270 + }, + { + "epoch": 24.812336339831248, + "grad_norm": 0.10284925252199173, + "learning_rate": 6.35835746103936e-06, + "loss": 0.0126, + "step": 85280 + }, + { + "epoch": 24.815245853942393, + "grad_norm": 0.10569377988576889, + "learning_rate": 6.354979037229389e-06, + "loss": 0.0125, + "step": 85290 + }, + { + "epoch": 24.818155368053535, + "grad_norm": 0.12465944141149521, + "learning_rate": 6.3516012699517065e-06, + "loss": 0.0124, + "step": 85300 + }, + { + "epoch": 24.821064882164677, + "grad_norm": 0.09027573466300964, + "learning_rate": 6.348224159462839e-06, + "loss": 0.0143, + "step": 85310 + }, + { + "epoch": 24.823974396275823, + "grad_norm": 0.1031322032213211, + "learning_rate": 6.3448477060192444e-06, + "loss": 0.0134, + "step": 85320 + }, + { + "epoch": 24.826883910386965, + "grad_norm": 0.10984901338815689, + "learning_rate": 6.341471909877353e-06, + "loss": 0.0127, + "step": 85330 + }, + { + "epoch": 24.82979342449811, + "grad_norm": 0.1103210374712944, + "learning_rate": 6.338096771293524e-06, + "loss": 0.0134, + "step": 85340 + }, + { + "epoch": 24.832702938609252, + "grad_norm": 0.10986834019422531, + "learning_rate": 6.334722290524081e-06, + "loss": 0.0117, + "step": 85350 + }, + { + "epoch": 24.835612452720397, + "grad_norm": 0.07670003175735474, + "learning_rate": 6.331348467825299e-06, + "loss": 0.0122, + "step": 85360 + }, + { + "epoch": 24.83852196683154, + "grad_norm": 0.11833784729242325, + "learning_rate": 6.327975303453388e-06, + "loss": 0.0125, + "step": 85370 + }, + { + "epoch": 24.84143148094268, + "grad_norm": 0.09652166813611984, + "learning_rate": 6.324602797664527e-06, + "loss": 0.0112, + "step": 85380 + }, + { + "epoch": 24.844340995053827, + "grad_norm": 0.10168454051017761, + "learning_rate": 6.321230950714828e-06, + "loss": 0.0124, + "step": 85390 + }, + { + "epoch": 24.84725050916497, + "grad_norm": 0.09591955691576004, + "learning_rate": 6.3178597628603666e-06, + "loss": 0.0102, + "step": 85400 + }, + { + "epoch": 24.850160023276114, + "grad_norm": 0.09037522226572037, + "learning_rate": 6.314489234357154e-06, + "loss": 0.0127, + "step": 85410 + }, + { + "epoch": 24.853069537387256, + "grad_norm": 0.07726765424013138, + "learning_rate": 6.311119365461163e-06, + "loss": 0.0115, + "step": 85420 + }, + { + "epoch": 24.8559790514984, + "grad_norm": 0.1147426962852478, + "learning_rate": 6.307750156428322e-06, + "loss": 0.0114, + "step": 85430 + }, + { + "epoch": 24.858888565609544, + "grad_norm": 0.1611948311328888, + "learning_rate": 6.304381607514493e-06, + "loss": 0.0122, + "step": 85440 + }, + { + "epoch": 24.861798079720685, + "grad_norm": 0.12793035805225372, + "learning_rate": 6.301013718975483e-06, + "loss": 0.0132, + "step": 85450 + }, + { + "epoch": 24.86470759383183, + "grad_norm": 0.16777218878269196, + "learning_rate": 6.297646491067085e-06, + "loss": 0.0123, + "step": 85460 + }, + { + "epoch": 24.867617107942973, + "grad_norm": 0.11690624058246613, + "learning_rate": 6.294279924045003e-06, + "loss": 0.0122, + "step": 85470 + }, + { + "epoch": 24.87052662205412, + "grad_norm": 0.10180002450942993, + "learning_rate": 6.290914018164905e-06, + "loss": 0.012, + "step": 85480 + }, + { + "epoch": 24.87343613616526, + "grad_norm": 0.10044217854738235, + "learning_rate": 6.287548773682411e-06, + "loss": 0.0137, + "step": 85490 + }, + { + "epoch": 24.876345650276402, + "grad_norm": 0.20322756469249725, + "learning_rate": 6.2841841908530955e-06, + "loss": 0.012, + "step": 85500 + }, + { + "epoch": 24.879255164387548, + "grad_norm": 0.058610253036022186, + "learning_rate": 6.280820269932472e-06, + "loss": 0.0106, + "step": 85510 + }, + { + "epoch": 24.88216467849869, + "grad_norm": 0.10222066938877106, + "learning_rate": 6.277457011176003e-06, + "loss": 0.0117, + "step": 85520 + }, + { + "epoch": 24.885074192609835, + "grad_norm": 0.09049922227859497, + "learning_rate": 6.274094414839109e-06, + "loss": 0.0107, + "step": 85530 + }, + { + "epoch": 24.887983706720977, + "grad_norm": 0.0967082679271698, + "learning_rate": 6.270732481177165e-06, + "loss": 0.0133, + "step": 85540 + }, + { + "epoch": 24.890893220832123, + "grad_norm": 0.0962807759642601, + "learning_rate": 6.267371210445473e-06, + "loss": 0.0105, + "step": 85550 + }, + { + "epoch": 24.893802734943264, + "grad_norm": 0.11145221441984177, + "learning_rate": 6.26401060289931e-06, + "loss": 0.0123, + "step": 85560 + }, + { + "epoch": 24.896712249054406, + "grad_norm": 0.10226771980524063, + "learning_rate": 6.260650658793892e-06, + "loss": 0.0125, + "step": 85570 + }, + { + "epoch": 24.899621763165552, + "grad_norm": 0.15415678918361664, + "learning_rate": 6.2572913783843845e-06, + "loss": 0.0136, + "step": 85580 + }, + { + "epoch": 24.902531277276694, + "grad_norm": 0.11573503911495209, + "learning_rate": 6.253932761925893e-06, + "loss": 0.0116, + "step": 85590 + }, + { + "epoch": 24.90544079138784, + "grad_norm": 0.1452203392982483, + "learning_rate": 6.250574809673489e-06, + "loss": 0.0102, + "step": 85600 + }, + { + "epoch": 24.90835030549898, + "grad_norm": 0.14916515350341797, + "learning_rate": 6.247217521882193e-06, + "loss": 0.0126, + "step": 85610 + }, + { + "epoch": 24.911259819610127, + "grad_norm": 0.10076643526554108, + "learning_rate": 6.243860898806956e-06, + "loss": 0.013, + "step": 85620 + }, + { + "epoch": 24.91416933372127, + "grad_norm": 0.10956788063049316, + "learning_rate": 6.240504940702706e-06, + "loss": 0.0124, + "step": 85630 + }, + { + "epoch": 24.91707884783241, + "grad_norm": 0.10100875794887543, + "learning_rate": 6.23714964782429e-06, + "loss": 0.0117, + "step": 85640 + }, + { + "epoch": 24.919988361943556, + "grad_norm": 0.15398548543453217, + "learning_rate": 6.2337950204265316e-06, + "loss": 0.0131, + "step": 85650 + }, + { + "epoch": 24.922897876054698, + "grad_norm": 0.10184542834758759, + "learning_rate": 6.230441058764193e-06, + "loss": 0.0134, + "step": 85660 + }, + { + "epoch": 24.925807390165843, + "grad_norm": 0.13151587545871735, + "learning_rate": 6.227087763091976e-06, + "loss": 0.0102, + "step": 85670 + }, + { + "epoch": 24.928716904276985, + "grad_norm": 0.10899757593870163, + "learning_rate": 6.223735133664553e-06, + "loss": 0.012, + "step": 85680 + }, + { + "epoch": 24.93162641838813, + "grad_norm": 0.18516609072685242, + "learning_rate": 6.220383170736524e-06, + "loss": 0.0098, + "step": 85690 + }, + { + "epoch": 24.934535932499273, + "grad_norm": 0.09418153017759323, + "learning_rate": 6.217031874562458e-06, + "loss": 0.0122, + "step": 85700 + }, + { + "epoch": 24.937445446610415, + "grad_norm": 0.09670817852020264, + "learning_rate": 6.213681245396853e-06, + "loss": 0.0114, + "step": 85710 + }, + { + "epoch": 24.94035496072156, + "grad_norm": 0.13291838765144348, + "learning_rate": 6.210331283494173e-06, + "loss": 0.011, + "step": 85720 + }, + { + "epoch": 24.943264474832702, + "grad_norm": 0.11332960426807404, + "learning_rate": 6.206981989108829e-06, + "loss": 0.0112, + "step": 85730 + }, + { + "epoch": 24.946173988943848, + "grad_norm": 0.09275561571121216, + "learning_rate": 6.203633362495175e-06, + "loss": 0.0126, + "step": 85740 + }, + { + "epoch": 24.94908350305499, + "grad_norm": 0.1205025166273117, + "learning_rate": 6.2002854039075105e-06, + "loss": 0.011, + "step": 85750 + }, + { + "epoch": 24.95199301716613, + "grad_norm": 0.13624876737594604, + "learning_rate": 6.196938113600098e-06, + "loss": 0.013, + "step": 85760 + }, + { + "epoch": 24.954902531277277, + "grad_norm": 0.13278517127037048, + "learning_rate": 6.1935914918271455e-06, + "loss": 0.0118, + "step": 85770 + }, + { + "epoch": 24.95781204538842, + "grad_norm": 0.09846209734678268, + "learning_rate": 6.190245538842794e-06, + "loss": 0.0105, + "step": 85780 + }, + { + "epoch": 24.960721559499564, + "grad_norm": 0.11793617159128189, + "learning_rate": 6.186900254901158e-06, + "loss": 0.013, + "step": 85790 + }, + { + "epoch": 24.963631073610706, + "grad_norm": 0.11861302703619003, + "learning_rate": 6.183555640256289e-06, + "loss": 0.0136, + "step": 85800 + }, + { + "epoch": 24.966540587721852, + "grad_norm": 0.09886655956506729, + "learning_rate": 6.180211695162188e-06, + "loss": 0.0117, + "step": 85810 + }, + { + "epoch": 24.969450101832994, + "grad_norm": 0.12021879851818085, + "learning_rate": 6.1768684198727955e-06, + "loss": 0.0154, + "step": 85820 + }, + { + "epoch": 24.972359615944136, + "grad_norm": 0.11797337234020233, + "learning_rate": 6.173525814642021e-06, + "loss": 0.0131, + "step": 85830 + }, + { + "epoch": 24.97526913005528, + "grad_norm": 0.11607497930526733, + "learning_rate": 6.1701838797237146e-06, + "loss": 0.013, + "step": 85840 + }, + { + "epoch": 24.978178644166423, + "grad_norm": 0.13021092116832733, + "learning_rate": 6.166842615371672e-06, + "loss": 0.0109, + "step": 85850 + }, + { + "epoch": 24.98108815827757, + "grad_norm": 0.13061492145061493, + "learning_rate": 6.163502021839628e-06, + "loss": 0.0099, + "step": 85860 + }, + { + "epoch": 24.98399767238871, + "grad_norm": 0.13041238486766815, + "learning_rate": 6.160162099381301e-06, + "loss": 0.0107, + "step": 85870 + }, + { + "epoch": 24.986907186499856, + "grad_norm": 0.06641504168510437, + "learning_rate": 6.156822848250324e-06, + "loss": 0.0109, + "step": 85880 + }, + { + "epoch": 24.989816700610998, + "grad_norm": 0.08063560724258423, + "learning_rate": 6.153484268700286e-06, + "loss": 0.0114, + "step": 85890 + }, + { + "epoch": 24.99272621472214, + "grad_norm": 0.09768492728471756, + "learning_rate": 6.150146360984738e-06, + "loss": 0.0119, + "step": 85900 + }, + { + "epoch": 24.995635728833285, + "grad_norm": 0.09927854686975479, + "learning_rate": 6.146809125357173e-06, + "loss": 0.0119, + "step": 85910 + }, + { + "epoch": 24.998545242944427, + "grad_norm": 0.0972532257437706, + "learning_rate": 6.1434725620710245e-06, + "loss": 0.0123, + "step": 85920 + }, + { + "epoch": 25.001454757055573, + "grad_norm": 0.11746270954608917, + "learning_rate": 6.140136671379692e-06, + "loss": 0.0107, + "step": 85930 + }, + { + "epoch": 25.004364271166715, + "grad_norm": 0.1110803559422493, + "learning_rate": 6.136801453536503e-06, + "loss": 0.0127, + "step": 85940 + }, + { + "epoch": 25.00727378527786, + "grad_norm": 0.08417966961860657, + "learning_rate": 6.133466908794758e-06, + "loss": 0.012, + "step": 85950 + }, + { + "epoch": 25.010183299389002, + "grad_norm": 0.10145967453718185, + "learning_rate": 6.130133037407683e-06, + "loss": 0.0129, + "step": 85960 + }, + { + "epoch": 25.013092813500144, + "grad_norm": 0.14896011352539062, + "learning_rate": 6.126799839628466e-06, + "loss": 0.0113, + "step": 85970 + }, + { + "epoch": 25.01600232761129, + "grad_norm": 0.12887533009052277, + "learning_rate": 6.123467315710249e-06, + "loss": 0.0119, + "step": 85980 + }, + { + "epoch": 25.01891184172243, + "grad_norm": 0.17322488129138947, + "learning_rate": 6.1201354659061045e-06, + "loss": 0.012, + "step": 85990 + }, + { + "epoch": 25.021821355833577, + "grad_norm": 0.19435974955558777, + "learning_rate": 6.116804290469075e-06, + "loss": 0.0142, + "step": 86000 + }, + { + "epoch": 25.02473086994472, + "grad_norm": 0.11961536854505539, + "learning_rate": 6.1134737896521295e-06, + "loss": 0.0125, + "step": 86010 + }, + { + "epoch": 25.027640384055864, + "grad_norm": 0.0866071805357933, + "learning_rate": 6.110143963708203e-06, + "loss": 0.0113, + "step": 86020 + }, + { + "epoch": 25.030549898167006, + "grad_norm": 0.0879877582192421, + "learning_rate": 6.106814812890182e-06, + "loss": 0.0104, + "step": 86030 + }, + { + "epoch": 25.033459412278148, + "grad_norm": 0.07898099720478058, + "learning_rate": 6.103486337450884e-06, + "loss": 0.0115, + "step": 86040 + }, + { + "epoch": 25.036368926389294, + "grad_norm": 0.12815001606941223, + "learning_rate": 6.100158537643081e-06, + "loss": 0.0116, + "step": 86050 + }, + { + "epoch": 25.039278440500436, + "grad_norm": 0.11045181006193161, + "learning_rate": 6.0968314137195055e-06, + "loss": 0.0125, + "step": 86060 + }, + { + "epoch": 25.04218795461158, + "grad_norm": 0.08950744569301605, + "learning_rate": 6.093504965932832e-06, + "loss": 0.0113, + "step": 86070 + }, + { + "epoch": 25.045097468722723, + "grad_norm": 0.09602786600589752, + "learning_rate": 6.090179194535674e-06, + "loss": 0.0125, + "step": 86080 + }, + { + "epoch": 25.04800698283387, + "grad_norm": 0.14716146886348724, + "learning_rate": 6.086854099780607e-06, + "loss": 0.0139, + "step": 86090 + }, + { + "epoch": 25.05091649694501, + "grad_norm": 0.08243688195943832, + "learning_rate": 6.083529681920155e-06, + "loss": 0.0141, + "step": 86100 + }, + { + "epoch": 25.053826011056152, + "grad_norm": 0.12707296013832092, + "learning_rate": 6.080205941206778e-06, + "loss": 0.0138, + "step": 86110 + }, + { + "epoch": 25.056735525167298, + "grad_norm": 0.09195158630609512, + "learning_rate": 6.076882877892891e-06, + "loss": 0.0125, + "step": 86120 + }, + { + "epoch": 25.05964503927844, + "grad_norm": 0.09495742619037628, + "learning_rate": 6.0735604922308605e-06, + "loss": 0.0131, + "step": 86130 + }, + { + "epoch": 25.062554553389585, + "grad_norm": 0.1219615563750267, + "learning_rate": 6.070238784473007e-06, + "loss": 0.0108, + "step": 86140 + }, + { + "epoch": 25.065464067500727, + "grad_norm": 0.11711146682500839, + "learning_rate": 6.066917754871587e-06, + "loss": 0.0103, + "step": 86150 + }, + { + "epoch": 25.06837358161187, + "grad_norm": 0.11943256109952927, + "learning_rate": 6.063597403678799e-06, + "loss": 0.0123, + "step": 86160 + }, + { + "epoch": 25.071283095723015, + "grad_norm": 0.11109407246112823, + "learning_rate": 6.060277731146825e-06, + "loss": 0.0133, + "step": 86170 + }, + { + "epoch": 25.074192609834157, + "grad_norm": 0.11558933556079865, + "learning_rate": 6.056958737527759e-06, + "loss": 0.012, + "step": 86180 + }, + { + "epoch": 25.077102123945302, + "grad_norm": 0.11845910549163818, + "learning_rate": 6.053640423073652e-06, + "loss": 0.0098, + "step": 86190 + }, + { + "epoch": 25.080011638056444, + "grad_norm": 0.13331936299800873, + "learning_rate": 6.050322788036515e-06, + "loss": 0.0109, + "step": 86200 + }, + { + "epoch": 25.08292115216759, + "grad_norm": 0.10883145779371262, + "learning_rate": 6.0470058326683055e-06, + "loss": 0.0141, + "step": 86210 + }, + { + "epoch": 25.08583066627873, + "grad_norm": 0.12354930490255356, + "learning_rate": 6.043689557220918e-06, + "loss": 0.0095, + "step": 86220 + }, + { + "epoch": 25.088740180389873, + "grad_norm": 0.12159726768732071, + "learning_rate": 6.040373961946196e-06, + "loss": 0.0125, + "step": 86230 + }, + { + "epoch": 25.09164969450102, + "grad_norm": 0.12891516089439392, + "learning_rate": 6.037059047095944e-06, + "loss": 0.0128, + "step": 86240 + }, + { + "epoch": 25.09455920861216, + "grad_norm": 0.08604128658771515, + "learning_rate": 6.033744812921911e-06, + "loss": 0.0118, + "step": 86250 + }, + { + "epoch": 25.097468722723306, + "grad_norm": 0.09768880158662796, + "learning_rate": 6.030431259675784e-06, + "loss": 0.0119, + "step": 86260 + }, + { + "epoch": 25.100378236834448, + "grad_norm": 0.09811443835496902, + "learning_rate": 6.027118387609208e-06, + "loss": 0.0125, + "step": 86270 + }, + { + "epoch": 25.103287750945594, + "grad_norm": 0.07936377078294754, + "learning_rate": 6.02380619697378e-06, + "loss": 0.0125, + "step": 86280 + }, + { + "epoch": 25.106197265056736, + "grad_norm": 0.08743051439523697, + "learning_rate": 6.020494688021028e-06, + "loss": 0.0107, + "step": 86290 + }, + { + "epoch": 25.109106779167877, + "grad_norm": 0.14093996584415436, + "learning_rate": 6.017183861002452e-06, + "loss": 0.0114, + "step": 86300 + }, + { + "epoch": 25.112016293279023, + "grad_norm": 0.11149803549051285, + "learning_rate": 6.013873716169473e-06, + "loss": 0.0134, + "step": 86310 + }, + { + "epoch": 25.114925807390165, + "grad_norm": 0.10292930155992508, + "learning_rate": 6.010564253773486e-06, + "loss": 0.0102, + "step": 86320 + }, + { + "epoch": 25.11783532150131, + "grad_norm": 0.11240074038505554, + "learning_rate": 6.0072554740658154e-06, + "loss": 0.0131, + "step": 86330 + }, + { + "epoch": 25.120744835612452, + "grad_norm": 0.14069773256778717, + "learning_rate": 6.003947377297751e-06, + "loss": 0.0121, + "step": 86340 + }, + { + "epoch": 25.123654349723598, + "grad_norm": 0.13305328786373138, + "learning_rate": 6.000639963720507e-06, + "loss": 0.0126, + "step": 86350 + }, + { + "epoch": 25.12656386383474, + "grad_norm": 0.1065889298915863, + "learning_rate": 5.997333233585268e-06, + "loss": 0.0134, + "step": 86360 + }, + { + "epoch": 25.12947337794588, + "grad_norm": 0.0846293717622757, + "learning_rate": 5.994027187143163e-06, + "loss": 0.0104, + "step": 86370 + }, + { + "epoch": 25.132382892057027, + "grad_norm": 0.15310800075531006, + "learning_rate": 5.990721824645254e-06, + "loss": 0.0126, + "step": 86380 + }, + { + "epoch": 25.13529240616817, + "grad_norm": 0.10960552841424942, + "learning_rate": 5.987417146342565e-06, + "loss": 0.0114, + "step": 86390 + }, + { + "epoch": 25.138201920279315, + "grad_norm": 0.0988236665725708, + "learning_rate": 5.9841131524860735e-06, + "loss": 0.0118, + "step": 86400 + }, + { + "epoch": 25.141111434390456, + "grad_norm": 0.1094384491443634, + "learning_rate": 5.980809843326685e-06, + "loss": 0.0124, + "step": 86410 + }, + { + "epoch": 25.144020948501602, + "grad_norm": 0.11255045235157013, + "learning_rate": 5.977507219115265e-06, + "loss": 0.012, + "step": 86420 + }, + { + "epoch": 25.146930462612744, + "grad_norm": 0.11158303171396255, + "learning_rate": 5.9742052801026265e-06, + "loss": 0.0126, + "step": 86430 + }, + { + "epoch": 25.149839976723886, + "grad_norm": 0.08343951404094696, + "learning_rate": 5.970904026539538e-06, + "loss": 0.0136, + "step": 86440 + }, + { + "epoch": 25.15274949083503, + "grad_norm": 0.12121092528104782, + "learning_rate": 5.967603458676703e-06, + "loss": 0.0116, + "step": 86450 + }, + { + "epoch": 25.155659004946173, + "grad_norm": 0.08739136904478073, + "learning_rate": 5.964303576764764e-06, + "loss": 0.0104, + "step": 86460 + }, + { + "epoch": 25.15856851905732, + "grad_norm": 0.10387146472930908, + "learning_rate": 5.961004381054351e-06, + "loss": 0.0146, + "step": 86470 + }, + { + "epoch": 25.16147803316846, + "grad_norm": 0.10584622621536255, + "learning_rate": 5.957705871796001e-06, + "loss": 0.0148, + "step": 86480 + }, + { + "epoch": 25.164387547279603, + "grad_norm": 0.11739082634449005, + "learning_rate": 5.954408049240211e-06, + "loss": 0.0126, + "step": 86490 + }, + { + "epoch": 25.167297061390748, + "grad_norm": 0.09067046642303467, + "learning_rate": 5.951110913637435e-06, + "loss": 0.0107, + "step": 86500 + }, + { + "epoch": 25.17020657550189, + "grad_norm": 0.101159006357193, + "learning_rate": 5.9478144652380725e-06, + "loss": 0.0128, + "step": 86510 + }, + { + "epoch": 25.173116089613035, + "grad_norm": 0.09772844612598419, + "learning_rate": 5.9445187042924615e-06, + "loss": 0.0123, + "step": 86520 + }, + { + "epoch": 25.176025603724177, + "grad_norm": 0.14358943700790405, + "learning_rate": 5.9412236310508896e-06, + "loss": 0.0137, + "step": 86530 + }, + { + "epoch": 25.178935117835323, + "grad_norm": 0.10577183216810226, + "learning_rate": 5.937929245763599e-06, + "loss": 0.0128, + "step": 86540 + }, + { + "epoch": 25.181844631946465, + "grad_norm": 0.13220950961112976, + "learning_rate": 5.934635548680784e-06, + "loss": 0.0134, + "step": 86550 + }, + { + "epoch": 25.184754146057607, + "grad_norm": 0.07032065838575363, + "learning_rate": 5.931342540052567e-06, + "loss": 0.012, + "step": 86560 + }, + { + "epoch": 25.187663660168752, + "grad_norm": 0.13498838245868683, + "learning_rate": 5.928050220129037e-06, + "loss": 0.0123, + "step": 86570 + }, + { + "epoch": 25.190573174279894, + "grad_norm": 0.12329194694757462, + "learning_rate": 5.9247585891602265e-06, + "loss": 0.0125, + "step": 86580 + }, + { + "epoch": 25.19348268839104, + "grad_norm": 0.11143108457326889, + "learning_rate": 5.921467647396112e-06, + "loss": 0.0127, + "step": 86590 + }, + { + "epoch": 25.19639220250218, + "grad_norm": 0.10464686155319214, + "learning_rate": 5.9181773950866114e-06, + "loss": 0.0117, + "step": 86600 + }, + { + "epoch": 25.199301716613327, + "grad_norm": 0.07959343492984772, + "learning_rate": 5.9148878324816016e-06, + "loss": 0.0129, + "step": 86610 + }, + { + "epoch": 25.20221123072447, + "grad_norm": 0.1128874197602272, + "learning_rate": 5.91159895983091e-06, + "loss": 0.0123, + "step": 86620 + }, + { + "epoch": 25.20512074483561, + "grad_norm": 0.11910408735275269, + "learning_rate": 5.908310777384292e-06, + "loss": 0.0114, + "step": 86630 + }, + { + "epoch": 25.208030258946756, + "grad_norm": 0.10409354418516159, + "learning_rate": 5.905023285391479e-06, + "loss": 0.0112, + "step": 86640 + }, + { + "epoch": 25.2109397730579, + "grad_norm": 0.12335251271724701, + "learning_rate": 5.901736484102119e-06, + "loss": 0.012, + "step": 86650 + }, + { + "epoch": 25.213849287169044, + "grad_norm": 0.11148657649755478, + "learning_rate": 5.898450373765829e-06, + "loss": 0.012, + "step": 86660 + }, + { + "epoch": 25.216758801280186, + "grad_norm": 0.10629821568727493, + "learning_rate": 5.895164954632176e-06, + "loss": 0.0096, + "step": 86670 + }, + { + "epoch": 25.21966831539133, + "grad_norm": 0.12796781957149506, + "learning_rate": 5.891880226950651e-06, + "loss": 0.0116, + "step": 86680 + }, + { + "epoch": 25.222577829502473, + "grad_norm": 0.09295260906219482, + "learning_rate": 5.88859619097072e-06, + "loss": 0.0106, + "step": 86690 + }, + { + "epoch": 25.225487343613615, + "grad_norm": 0.1172393187880516, + "learning_rate": 5.885312846941772e-06, + "loss": 0.0122, + "step": 86700 + }, + { + "epoch": 25.22839685772476, + "grad_norm": 0.11674809455871582, + "learning_rate": 5.882030195113166e-06, + "loss": 0.0116, + "step": 86710 + }, + { + "epoch": 25.231306371835903, + "grad_norm": 0.0941578820347786, + "learning_rate": 5.87874823573419e-06, + "loss": 0.0109, + "step": 86720 + }, + { + "epoch": 25.234215885947048, + "grad_norm": 0.10578205436468124, + "learning_rate": 5.8754669690540905e-06, + "loss": 0.0127, + "step": 86730 + }, + { + "epoch": 25.23712540005819, + "grad_norm": 0.0962415337562561, + "learning_rate": 5.872186395322063e-06, + "loss": 0.011, + "step": 86740 + }, + { + "epoch": 25.240034914169335, + "grad_norm": 0.11222007870674133, + "learning_rate": 5.8689065147872404e-06, + "loss": 0.0113, + "step": 86750 + }, + { + "epoch": 25.242944428280477, + "grad_norm": 0.10162713378667831, + "learning_rate": 5.865627327698698e-06, + "loss": 0.0112, + "step": 86760 + }, + { + "epoch": 25.24585394239162, + "grad_norm": 0.1168115884065628, + "learning_rate": 5.862348834305489e-06, + "loss": 0.0131, + "step": 86770 + }, + { + "epoch": 25.248763456502765, + "grad_norm": 0.07431214302778244, + "learning_rate": 5.859071034856582e-06, + "loss": 0.0133, + "step": 86780 + }, + { + "epoch": 25.251672970613907, + "grad_norm": 0.1111876592040062, + "learning_rate": 5.8557939296009e-06, + "loss": 0.0121, + "step": 86790 + }, + { + "epoch": 25.254582484725052, + "grad_norm": 0.09611357748508453, + "learning_rate": 5.852517518787323e-06, + "loss": 0.0113, + "step": 86800 + }, + { + "epoch": 25.257491998836194, + "grad_norm": 0.09273262321949005, + "learning_rate": 5.8492418026646766e-06, + "loss": 0.0117, + "step": 86810 + }, + { + "epoch": 25.26040151294734, + "grad_norm": 0.1644279956817627, + "learning_rate": 5.845966781481725e-06, + "loss": 0.0132, + "step": 86820 + }, + { + "epoch": 25.26331102705848, + "grad_norm": 0.1441599726676941, + "learning_rate": 5.842692455487181e-06, + "loss": 0.0136, + "step": 86830 + }, + { + "epoch": 25.266220541169623, + "grad_norm": 0.2767721116542816, + "learning_rate": 5.839418824929713e-06, + "loss": 0.012, + "step": 86840 + }, + { + "epoch": 25.26913005528077, + "grad_norm": 0.1296311616897583, + "learning_rate": 5.836145890057933e-06, + "loss": 0.0134, + "step": 86850 + }, + { + "epoch": 25.27203956939191, + "grad_norm": 0.0798201784491539, + "learning_rate": 5.832873651120394e-06, + "loss": 0.0106, + "step": 86860 + }, + { + "epoch": 25.274949083503056, + "grad_norm": 0.1196063980460167, + "learning_rate": 5.829602108365602e-06, + "loss": 0.0127, + "step": 86870 + }, + { + "epoch": 25.2778585976142, + "grad_norm": 0.08389527350664139, + "learning_rate": 5.826331262042015e-06, + "loss": 0.0107, + "step": 86880 + }, + { + "epoch": 25.28076811172534, + "grad_norm": 0.08662257343530655, + "learning_rate": 5.823061112398031e-06, + "loss": 0.0115, + "step": 86890 + }, + { + "epoch": 25.283677625836486, + "grad_norm": 0.08979322016239166, + "learning_rate": 5.8197916596819844e-06, + "loss": 0.0109, + "step": 86900 + }, + { + "epoch": 25.286587139947628, + "grad_norm": 0.1421636939048767, + "learning_rate": 5.816522904142179e-06, + "loss": 0.0138, + "step": 86910 + }, + { + "epoch": 25.289496654058773, + "grad_norm": 0.13639117777347565, + "learning_rate": 5.8132548460268575e-06, + "loss": 0.0125, + "step": 86920 + }, + { + "epoch": 25.292406168169915, + "grad_norm": 0.1142483577132225, + "learning_rate": 5.8099874855841974e-06, + "loss": 0.0115, + "step": 86930 + }, + { + "epoch": 25.29531568228106, + "grad_norm": 0.09981069713830948, + "learning_rate": 5.8067208230623455e-06, + "loss": 0.0133, + "step": 86940 + }, + { + "epoch": 25.298225196392202, + "grad_norm": 0.17936305701732635, + "learning_rate": 5.803454858709371e-06, + "loss": 0.0111, + "step": 86950 + }, + { + "epoch": 25.301134710503344, + "grad_norm": 0.12150456011295319, + "learning_rate": 5.8001895927733135e-06, + "loss": 0.0121, + "step": 86960 + }, + { + "epoch": 25.30404422461449, + "grad_norm": 0.14500518143177032, + "learning_rate": 5.796925025502138e-06, + "loss": 0.0149, + "step": 86970 + }, + { + "epoch": 25.306953738725632, + "grad_norm": 0.12600308656692505, + "learning_rate": 5.793661157143772e-06, + "loss": 0.0113, + "step": 86980 + }, + { + "epoch": 25.309863252836777, + "grad_norm": 0.12201393395662308, + "learning_rate": 5.790397987946088e-06, + "loss": 0.0122, + "step": 86990 + }, + { + "epoch": 25.31277276694792, + "grad_norm": 0.08881831914186478, + "learning_rate": 5.787135518156895e-06, + "loss": 0.0143, + "step": 87000 + }, + { + "epoch": 25.315682281059065, + "grad_norm": 0.10068480670452118, + "learning_rate": 5.783873748023966e-06, + "loss": 0.013, + "step": 87010 + }, + { + "epoch": 25.318591795170207, + "grad_norm": 0.08728103339672089, + "learning_rate": 5.7806126777949975e-06, + "loss": 0.0112, + "step": 87020 + }, + { + "epoch": 25.32150130928135, + "grad_norm": 0.10652013123035431, + "learning_rate": 5.7773523077176544e-06, + "loss": 0.0114, + "step": 87030 + }, + { + "epoch": 25.324410823392494, + "grad_norm": 0.12219797819852829, + "learning_rate": 5.774092638039543e-06, + "loss": 0.0132, + "step": 87040 + }, + { + "epoch": 25.327320337503636, + "grad_norm": 0.08915212750434875, + "learning_rate": 5.770833669008213e-06, + "loss": 0.0102, + "step": 87050 + }, + { + "epoch": 25.33022985161478, + "grad_norm": 0.12581028044223785, + "learning_rate": 5.7675754008711515e-06, + "loss": 0.0138, + "step": 87060 + }, + { + "epoch": 25.333139365725923, + "grad_norm": 0.12684214115142822, + "learning_rate": 5.76431783387581e-06, + "loss": 0.0136, + "step": 87070 + }, + { + "epoch": 25.33604887983707, + "grad_norm": 0.08699925988912582, + "learning_rate": 5.761060968269583e-06, + "loss": 0.0126, + "step": 87080 + }, + { + "epoch": 25.33895839394821, + "grad_norm": 0.11143616586923599, + "learning_rate": 5.757804804299801e-06, + "loss": 0.0092, + "step": 87090 + }, + { + "epoch": 25.341867908059353, + "grad_norm": 0.13559377193450928, + "learning_rate": 5.754549342213749e-06, + "loss": 0.0154, + "step": 87100 + }, + { + "epoch": 25.3447774221705, + "grad_norm": 0.0945858359336853, + "learning_rate": 5.751294582258665e-06, + "loss": 0.0124, + "step": 87110 + }, + { + "epoch": 25.34768693628164, + "grad_norm": 0.09698645025491714, + "learning_rate": 5.748040524681722e-06, + "loss": 0.0134, + "step": 87120 + }, + { + "epoch": 25.350596450392786, + "grad_norm": 0.1870017945766449, + "learning_rate": 5.744787169730037e-06, + "loss": 0.0111, + "step": 87130 + }, + { + "epoch": 25.353505964503928, + "grad_norm": 0.13926786184310913, + "learning_rate": 5.7415345176506905e-06, + "loss": 0.0107, + "step": 87140 + }, + { + "epoch": 25.35641547861507, + "grad_norm": 0.07746720314025879, + "learning_rate": 5.738282568690699e-06, + "loss": 0.0119, + "step": 87150 + }, + { + "epoch": 25.359324992726215, + "grad_norm": 0.11385714262723923, + "learning_rate": 5.7350313230970255e-06, + "loss": 0.0105, + "step": 87160 + }, + { + "epoch": 25.362234506837357, + "grad_norm": 0.07208363711833954, + "learning_rate": 5.731780781116569e-06, + "loss": 0.0111, + "step": 87170 + }, + { + "epoch": 25.365144020948502, + "grad_norm": 0.11877432465553284, + "learning_rate": 5.728530942996207e-06, + "loss": 0.0131, + "step": 87180 + }, + { + "epoch": 25.368053535059644, + "grad_norm": 0.10301011055707932, + "learning_rate": 5.725281808982735e-06, + "loss": 0.0137, + "step": 87190 + }, + { + "epoch": 25.37096304917079, + "grad_norm": 0.08781221508979797, + "learning_rate": 5.722033379322895e-06, + "loss": 0.0118, + "step": 87200 + }, + { + "epoch": 25.37387256328193, + "grad_norm": 0.26018327474594116, + "learning_rate": 5.718785654263392e-06, + "loss": 0.0123, + "step": 87210 + }, + { + "epoch": 25.376782077393074, + "grad_norm": 0.11429275572299957, + "learning_rate": 5.715538634050872e-06, + "loss": 0.0117, + "step": 87220 + }, + { + "epoch": 25.37969159150422, + "grad_norm": 0.16038130223751068, + "learning_rate": 5.712292318931916e-06, + "loss": 0.0125, + "step": 87230 + }, + { + "epoch": 25.38260110561536, + "grad_norm": 0.11496246606111526, + "learning_rate": 5.70904670915307e-06, + "loss": 0.0124, + "step": 87240 + }, + { + "epoch": 25.385510619726507, + "grad_norm": 0.09996509552001953, + "learning_rate": 5.705801804960805e-06, + "loss": 0.0128, + "step": 87250 + }, + { + "epoch": 25.38842013383765, + "grad_norm": 0.11457006633281708, + "learning_rate": 5.702557606601564e-06, + "loss": 0.0137, + "step": 87260 + }, + { + "epoch": 25.391329647948794, + "grad_norm": 0.12397462874650955, + "learning_rate": 5.699314114321709e-06, + "loss": 0.0136, + "step": 87270 + }, + { + "epoch": 25.394239162059936, + "grad_norm": 0.11455778777599335, + "learning_rate": 5.696071328367568e-06, + "loss": 0.0147, + "step": 87280 + }, + { + "epoch": 25.397148676171078, + "grad_norm": 0.11188450455665588, + "learning_rate": 5.692829248985414e-06, + "loss": 0.0121, + "step": 87290 + }, + { + "epoch": 25.400058190282223, + "grad_norm": 0.1751265972852707, + "learning_rate": 5.689587876421451e-06, + "loss": 0.0113, + "step": 87300 + }, + { + "epoch": 25.402967704393365, + "grad_norm": 0.09795019775629044, + "learning_rate": 5.6863472109218506e-06, + "loss": 0.0138, + "step": 87310 + }, + { + "epoch": 25.40587721850451, + "grad_norm": 0.12966565787792206, + "learning_rate": 5.683107252732709e-06, + "loss": 0.0124, + "step": 87320 + }, + { + "epoch": 25.408786732615653, + "grad_norm": 0.12447944283485413, + "learning_rate": 5.679868002100092e-06, + "loss": 0.0136, + "step": 87330 + }, + { + "epoch": 25.411696246726798, + "grad_norm": 0.0785774365067482, + "learning_rate": 5.676629459269987e-06, + "loss": 0.0104, + "step": 87340 + }, + { + "epoch": 25.41460576083794, + "grad_norm": 0.09878047555685043, + "learning_rate": 5.67339162448835e-06, + "loss": 0.0132, + "step": 87350 + }, + { + "epoch": 25.417515274949082, + "grad_norm": 0.1129290759563446, + "learning_rate": 5.670154498001064e-06, + "loss": 0.0112, + "step": 87360 + }, + { + "epoch": 25.420424789060228, + "grad_norm": 0.1084652841091156, + "learning_rate": 5.666918080053974e-06, + "loss": 0.0138, + "step": 87370 + }, + { + "epoch": 25.42333430317137, + "grad_norm": 0.10602705925703049, + "learning_rate": 5.663682370892867e-06, + "loss": 0.0112, + "step": 87380 + }, + { + "epoch": 25.426243817282515, + "grad_norm": 0.1279289871454239, + "learning_rate": 5.660447370763464e-06, + "loss": 0.0134, + "step": 87390 + }, + { + "epoch": 25.429153331393657, + "grad_norm": 0.11082940548658371, + "learning_rate": 5.657213079911447e-06, + "loss": 0.0122, + "step": 87400 + }, + { + "epoch": 25.432062845504802, + "grad_norm": 0.12318835407495499, + "learning_rate": 5.653979498582445e-06, + "loss": 0.0139, + "step": 87410 + }, + { + "epoch": 25.434972359615944, + "grad_norm": 0.10859766602516174, + "learning_rate": 5.650746627022022e-06, + "loss": 0.0114, + "step": 87420 + }, + { + "epoch": 25.437881873727086, + "grad_norm": 0.08902312070131302, + "learning_rate": 5.647514465475689e-06, + "loss": 0.0128, + "step": 87430 + }, + { + "epoch": 25.44079138783823, + "grad_norm": 0.14400836825370789, + "learning_rate": 5.6442830141889094e-06, + "loss": 0.0122, + "step": 87440 + }, + { + "epoch": 25.443700901949374, + "grad_norm": 0.14861242473125458, + "learning_rate": 5.641052273407098e-06, + "loss": 0.0124, + "step": 87450 + }, + { + "epoch": 25.44661041606052, + "grad_norm": 0.122688427567482, + "learning_rate": 5.637822243375604e-06, + "loss": 0.0159, + "step": 87460 + }, + { + "epoch": 25.44951993017166, + "grad_norm": 0.10872464627027512, + "learning_rate": 5.634592924339714e-06, + "loss": 0.0116, + "step": 87470 + }, + { + "epoch": 25.452429444282807, + "grad_norm": 0.07674426585435867, + "learning_rate": 5.631364316544696e-06, + "loss": 0.012, + "step": 87480 + }, + { + "epoch": 25.45533895839395, + "grad_norm": 0.1386343538761139, + "learning_rate": 5.628136420235731e-06, + "loss": 0.01, + "step": 87490 + }, + { + "epoch": 25.45824847250509, + "grad_norm": 0.07711838185787201, + "learning_rate": 5.62490923565795e-06, + "loss": 0.0128, + "step": 87500 + }, + { + "epoch": 25.461157986616236, + "grad_norm": 0.09883510321378708, + "learning_rate": 5.621682763056443e-06, + "loss": 0.0111, + "step": 87510 + }, + { + "epoch": 25.464067500727378, + "grad_norm": 0.13884173333644867, + "learning_rate": 5.6184570026762425e-06, + "loss": 0.0144, + "step": 87520 + }, + { + "epoch": 25.466977014838523, + "grad_norm": 0.11150261759757996, + "learning_rate": 5.615231954762322e-06, + "loss": 0.0124, + "step": 87530 + }, + { + "epoch": 25.469886528949665, + "grad_norm": 0.12661266326904297, + "learning_rate": 5.612007619559594e-06, + "loss": 0.0101, + "step": 87540 + }, + { + "epoch": 25.472796043060807, + "grad_norm": 0.09870382398366928, + "learning_rate": 5.608783997312932e-06, + "loss": 0.0123, + "step": 87550 + }, + { + "epoch": 25.475705557171953, + "grad_norm": 0.09002383798360825, + "learning_rate": 5.605561088267154e-06, + "loss": 0.0116, + "step": 87560 + }, + { + "epoch": 25.478615071283095, + "grad_norm": 0.10871091485023499, + "learning_rate": 5.602338892667007e-06, + "loss": 0.0113, + "step": 87570 + }, + { + "epoch": 25.48152458539424, + "grad_norm": 0.11083795130252838, + "learning_rate": 5.599117410757203e-06, + "loss": 0.0113, + "step": 87580 + }, + { + "epoch": 25.484434099505382, + "grad_norm": 0.07435361295938492, + "learning_rate": 5.595896642782397e-06, + "loss": 0.0123, + "step": 87590 + }, + { + "epoch": 25.487343613616527, + "grad_norm": 0.15081092715263367, + "learning_rate": 5.592676588987174e-06, + "loss": 0.0137, + "step": 87600 + }, + { + "epoch": 25.49025312772767, + "grad_norm": 0.17939072847366333, + "learning_rate": 5.589457249616086e-06, + "loss": 0.0113, + "step": 87610 + }, + { + "epoch": 25.49316264183881, + "grad_norm": 0.1547721028327942, + "learning_rate": 5.58623862491361e-06, + "loss": 0.0138, + "step": 87620 + }, + { + "epoch": 25.496072155949957, + "grad_norm": 0.1277560591697693, + "learning_rate": 5.583020715124191e-06, + "loss": 0.0126, + "step": 87630 + }, + { + "epoch": 25.4989816700611, + "grad_norm": 0.13131898641586304, + "learning_rate": 5.579803520492197e-06, + "loss": 0.0127, + "step": 87640 + }, + { + "epoch": 25.501891184172244, + "grad_norm": 0.10803813487291336, + "learning_rate": 5.576587041261962e-06, + "loss": 0.0122, + "step": 87650 + }, + { + "epoch": 25.504800698283386, + "grad_norm": 0.09478063881397247, + "learning_rate": 5.573371277677748e-06, + "loss": 0.0135, + "step": 87660 + }, + { + "epoch": 25.50771021239453, + "grad_norm": 0.08231937140226364, + "learning_rate": 5.5701562299837745e-06, + "loss": 0.012, + "step": 87670 + }, + { + "epoch": 25.510619726505674, + "grad_norm": 0.11144640296697617, + "learning_rate": 5.566941898424208e-06, + "loss": 0.011, + "step": 87680 + }, + { + "epoch": 25.513529240616815, + "grad_norm": 0.11873196810483932, + "learning_rate": 5.5637282832431476e-06, + "loss": 0.0112, + "step": 87690 + }, + { + "epoch": 25.51643875472796, + "grad_norm": 0.07882092893123627, + "learning_rate": 5.56051538468465e-06, + "loss": 0.0115, + "step": 87700 + }, + { + "epoch": 25.519348268839103, + "grad_norm": 0.10871482640504837, + "learning_rate": 5.557303202992719e-06, + "loss": 0.0112, + "step": 87710 + }, + { + "epoch": 25.52225778295025, + "grad_norm": 0.10573923587799072, + "learning_rate": 5.554091738411292e-06, + "loss": 0.011, + "step": 87720 + }, + { + "epoch": 25.52516729706139, + "grad_norm": 0.07802636921405792, + "learning_rate": 5.550880991184256e-06, + "loss": 0.0117, + "step": 87730 + }, + { + "epoch": 25.528076811172536, + "grad_norm": 0.11789896339178085, + "learning_rate": 5.5476709615554514e-06, + "loss": 0.0122, + "step": 87740 + }, + { + "epoch": 25.530986325283678, + "grad_norm": 0.11713097244501114, + "learning_rate": 5.5444616497686605e-06, + "loss": 0.0098, + "step": 87750 + }, + { + "epoch": 25.53389583939482, + "grad_norm": 0.09881651401519775, + "learning_rate": 5.5412530560676065e-06, + "loss": 0.0124, + "step": 87760 + }, + { + "epoch": 25.536805353505965, + "grad_norm": 0.10137613862752914, + "learning_rate": 5.5380451806959495e-06, + "loss": 0.0127, + "step": 87770 + }, + { + "epoch": 25.539714867617107, + "grad_norm": 0.11559659987688065, + "learning_rate": 5.534838023897329e-06, + "loss": 0.0116, + "step": 87780 + }, + { + "epoch": 25.542624381728253, + "grad_norm": 0.08491960912942886, + "learning_rate": 5.531631585915296e-06, + "loss": 0.0101, + "step": 87790 + }, + { + "epoch": 25.545533895839394, + "grad_norm": 0.11502115428447723, + "learning_rate": 5.528425866993353e-06, + "loss": 0.0122, + "step": 87800 + }, + { + "epoch": 25.54844340995054, + "grad_norm": 0.06962329149246216, + "learning_rate": 5.525220867374957e-06, + "loss": 0.0117, + "step": 87810 + }, + { + "epoch": 25.551352924061682, + "grad_norm": 0.1086135283112526, + "learning_rate": 5.5220165873035156e-06, + "loss": 0.0119, + "step": 87820 + }, + { + "epoch": 25.554262438172824, + "grad_norm": 0.10104059427976608, + "learning_rate": 5.518813027022363e-06, + "loss": 0.0131, + "step": 87830 + }, + { + "epoch": 25.55717195228397, + "grad_norm": 0.1369713544845581, + "learning_rate": 5.515610186774787e-06, + "loss": 0.0101, + "step": 87840 + }, + { + "epoch": 25.56008146639511, + "grad_norm": 0.11611921340227127, + "learning_rate": 5.512408066804025e-06, + "loss": 0.0125, + "step": 87850 + }, + { + "epoch": 25.562990980506257, + "grad_norm": 0.08678068965673447, + "learning_rate": 5.509206667353264e-06, + "loss": 0.0108, + "step": 87860 + }, + { + "epoch": 25.5659004946174, + "grad_norm": 0.09598438441753387, + "learning_rate": 5.506005988665617e-06, + "loss": 0.0126, + "step": 87870 + }, + { + "epoch": 25.56881000872854, + "grad_norm": 0.09993021935224533, + "learning_rate": 5.502806030984159e-06, + "loss": 0.011, + "step": 87880 + }, + { + "epoch": 25.571719522839686, + "grad_norm": 0.12174860388040543, + "learning_rate": 5.499606794551913e-06, + "loss": 0.0112, + "step": 87890 + }, + { + "epoch": 25.574629036950828, + "grad_norm": 0.11917684227228165, + "learning_rate": 5.496408279611833e-06, + "loss": 0.0118, + "step": 87900 + }, + { + "epoch": 25.577538551061973, + "grad_norm": 0.11314687132835388, + "learning_rate": 5.493210486406823e-06, + "loss": 0.0101, + "step": 87910 + }, + { + "epoch": 25.580448065173115, + "grad_norm": 0.12913261353969574, + "learning_rate": 5.490013415179735e-06, + "loss": 0.0116, + "step": 87920 + }, + { + "epoch": 25.58335757928426, + "grad_norm": 0.11864857375621796, + "learning_rate": 5.486817066173374e-06, + "loss": 0.0104, + "step": 87930 + }, + { + "epoch": 25.586267093395403, + "grad_norm": 0.10822846740484238, + "learning_rate": 5.483621439630469e-06, + "loss": 0.0124, + "step": 87940 + }, + { + "epoch": 25.589176607506545, + "grad_norm": 0.11641058325767517, + "learning_rate": 5.480426535793719e-06, + "loss": 0.0121, + "step": 87950 + }, + { + "epoch": 25.59208612161769, + "grad_norm": 0.10689792037010193, + "learning_rate": 5.477232354905746e-06, + "loss": 0.011, + "step": 87960 + }, + { + "epoch": 25.594995635728832, + "grad_norm": 0.11325965821743011, + "learning_rate": 5.474038897209129e-06, + "loss": 0.0109, + "step": 87970 + }, + { + "epoch": 25.597905149839978, + "grad_norm": 0.09292846918106079, + "learning_rate": 5.470846162946398e-06, + "loss": 0.0118, + "step": 87980 + }, + { + "epoch": 25.60081466395112, + "grad_norm": 0.08531259000301361, + "learning_rate": 5.4676541523600075e-06, + "loss": 0.0101, + "step": 87990 + }, + { + "epoch": 25.603724178062265, + "grad_norm": 0.08686395734548569, + "learning_rate": 5.464462865692382e-06, + "loss": 0.0113, + "step": 88000 + }, + { + "epoch": 25.606633692173407, + "grad_norm": 0.11074709892272949, + "learning_rate": 5.461272303185868e-06, + "loss": 0.0131, + "step": 88010 + }, + { + "epoch": 25.60954320628455, + "grad_norm": 0.11511655896902084, + "learning_rate": 5.458082465082778e-06, + "loss": 0.0123, + "step": 88020 + }, + { + "epoch": 25.612452720395694, + "grad_norm": 0.10177662968635559, + "learning_rate": 5.454893351625348e-06, + "loss": 0.0115, + "step": 88030 + }, + { + "epoch": 25.615362234506836, + "grad_norm": 0.06893067806959152, + "learning_rate": 5.4517049630557756e-06, + "loss": 0.011, + "step": 88040 + }, + { + "epoch": 25.618271748617982, + "grad_norm": 0.0794714167714119, + "learning_rate": 5.448517299616203e-06, + "loss": 0.0107, + "step": 88050 + }, + { + "epoch": 25.621181262729124, + "grad_norm": 0.08444253355264664, + "learning_rate": 5.445330361548707e-06, + "loss": 0.0128, + "step": 88060 + }, + { + "epoch": 25.62409077684027, + "grad_norm": 0.1005539670586586, + "learning_rate": 5.442144149095304e-06, + "loss": 0.0122, + "step": 88070 + }, + { + "epoch": 25.62700029095141, + "grad_norm": 0.10637499392032623, + "learning_rate": 5.438958662497988e-06, + "loss": 0.015, + "step": 88080 + }, + { + "epoch": 25.629909805062553, + "grad_norm": 0.11277727037668228, + "learning_rate": 5.435773901998663e-06, + "loss": 0.0117, + "step": 88090 + }, + { + "epoch": 25.6328193191737, + "grad_norm": 0.08572866022586823, + "learning_rate": 5.432589867839189e-06, + "loss": 0.0131, + "step": 88100 + }, + { + "epoch": 25.63572883328484, + "grad_norm": 0.15232455730438232, + "learning_rate": 5.429406560261372e-06, + "loss": 0.0142, + "step": 88110 + }, + { + "epoch": 25.638638347395986, + "grad_norm": 0.0767337903380394, + "learning_rate": 5.4262239795069734e-06, + "loss": 0.0114, + "step": 88120 + }, + { + "epoch": 25.641547861507128, + "grad_norm": 0.12770430743694305, + "learning_rate": 5.423042125817681e-06, + "loss": 0.012, + "step": 88130 + }, + { + "epoch": 25.644457375618273, + "grad_norm": 0.09571942687034607, + "learning_rate": 5.419860999435133e-06, + "loss": 0.0109, + "step": 88140 + }, + { + "epoch": 25.647366889729415, + "grad_norm": 0.09836230427026749, + "learning_rate": 5.4166806006009175e-06, + "loss": 0.0115, + "step": 88150 + }, + { + "epoch": 25.650276403840557, + "grad_norm": 0.103660948574543, + "learning_rate": 5.413500929556572e-06, + "loss": 0.0104, + "step": 88160 + }, + { + "epoch": 25.653185917951703, + "grad_norm": 0.08697967976331711, + "learning_rate": 5.410321986543565e-06, + "loss": 0.0128, + "step": 88170 + }, + { + "epoch": 25.656095432062845, + "grad_norm": 0.14189639687538147, + "learning_rate": 5.407143771803305e-06, + "loss": 0.0095, + "step": 88180 + }, + { + "epoch": 25.65900494617399, + "grad_norm": 0.18440014123916626, + "learning_rate": 5.403966285577181e-06, + "loss": 0.0109, + "step": 88190 + }, + { + "epoch": 25.661914460285132, + "grad_norm": 0.1162438839673996, + "learning_rate": 5.4007895281064874e-06, + "loss": 0.0117, + "step": 88200 + }, + { + "epoch": 25.664823974396278, + "grad_norm": 0.09603282809257507, + "learning_rate": 5.397613499632476e-06, + "loss": 0.0124, + "step": 88210 + }, + { + "epoch": 25.66773348850742, + "grad_norm": 0.11100269109010696, + "learning_rate": 5.39443820039635e-06, + "loss": 0.0115, + "step": 88220 + }, + { + "epoch": 25.67064300261856, + "grad_norm": 0.11508617550134659, + "learning_rate": 5.391263630639255e-06, + "loss": 0.0145, + "step": 88230 + }, + { + "epoch": 25.673552516729707, + "grad_norm": 0.08719713240861893, + "learning_rate": 5.38808979060227e-06, + "loss": 0.0108, + "step": 88240 + }, + { + "epoch": 25.67646203084085, + "grad_norm": 0.12943428754806519, + "learning_rate": 5.384916680526438e-06, + "loss": 0.0127, + "step": 88250 + }, + { + "epoch": 25.679371544951994, + "grad_norm": 0.08845295757055283, + "learning_rate": 5.381744300652724e-06, + "loss": 0.0101, + "step": 88260 + }, + { + "epoch": 25.682281059063136, + "grad_norm": 0.11877479404211044, + "learning_rate": 5.37857265122206e-06, + "loss": 0.0116, + "step": 88270 + }, + { + "epoch": 25.68519057317428, + "grad_norm": 0.08659867942333221, + "learning_rate": 5.375401732475303e-06, + "loss": 0.0113, + "step": 88280 + }, + { + "epoch": 25.688100087285424, + "grad_norm": 0.08103295415639877, + "learning_rate": 5.372231544653269e-06, + "loss": 0.0132, + "step": 88290 + }, + { + "epoch": 25.691009601396566, + "grad_norm": 0.11643160134553909, + "learning_rate": 5.369062087996716e-06, + "loss": 0.014, + "step": 88300 + }, + { + "epoch": 25.69391911550771, + "grad_norm": 0.11680260300636292, + "learning_rate": 5.3658933627463334e-06, + "loss": 0.0111, + "step": 88310 + }, + { + "epoch": 25.696828629618853, + "grad_norm": 0.11752639710903168, + "learning_rate": 5.362725369142775e-06, + "loss": 0.0128, + "step": 88320 + }, + { + "epoch": 25.69973814373, + "grad_norm": 0.10498029738664627, + "learning_rate": 5.359558107426618e-06, + "loss": 0.0099, + "step": 88330 + }, + { + "epoch": 25.70264765784114, + "grad_norm": 0.07952028512954712, + "learning_rate": 5.3563915778384046e-06, + "loss": 0.0123, + "step": 88340 + }, + { + "epoch": 25.705557171952282, + "grad_norm": 0.13650281727313995, + "learning_rate": 5.353225780618612e-06, + "loss": 0.0141, + "step": 88350 + }, + { + "epoch": 25.708466686063428, + "grad_norm": 0.10780907422304153, + "learning_rate": 5.350060716007657e-06, + "loss": 0.0111, + "step": 88360 + }, + { + "epoch": 25.71137620017457, + "grad_norm": 0.0968942940235138, + "learning_rate": 5.346896384245902e-06, + "loss": 0.0115, + "step": 88370 + }, + { + "epoch": 25.714285714285715, + "grad_norm": 0.15827828645706177, + "learning_rate": 5.3437327855736615e-06, + "loss": 0.0109, + "step": 88380 + }, + { + "epoch": 25.717195228396857, + "grad_norm": 0.10371927917003632, + "learning_rate": 5.340569920231196e-06, + "loss": 0.0123, + "step": 88390 + }, + { + "epoch": 25.720104742508003, + "grad_norm": 0.1169683113694191, + "learning_rate": 5.337407788458694e-06, + "loss": 0.0139, + "step": 88400 + }, + { + "epoch": 25.723014256619145, + "grad_norm": 0.10406654328107834, + "learning_rate": 5.334246390496302e-06, + "loss": 0.0109, + "step": 88410 + }, + { + "epoch": 25.725923770730287, + "grad_norm": 0.0923481360077858, + "learning_rate": 5.331085726584114e-06, + "loss": 0.0115, + "step": 88420 + }, + { + "epoch": 25.728833284841432, + "grad_norm": 0.14993013441562653, + "learning_rate": 5.327925796962155e-06, + "loss": 0.0123, + "step": 88430 + }, + { + "epoch": 25.731742798952574, + "grad_norm": 0.13100889325141907, + "learning_rate": 5.324766601870396e-06, + "loss": 0.0129, + "step": 88440 + }, + { + "epoch": 25.73465231306372, + "grad_norm": 0.09301172941923141, + "learning_rate": 5.321608141548763e-06, + "loss": 0.012, + "step": 88450 + }, + { + "epoch": 25.73756182717486, + "grad_norm": 0.09636744856834412, + "learning_rate": 5.318450416237125e-06, + "loss": 0.0116, + "step": 88460 + }, + { + "epoch": 25.740471341286007, + "grad_norm": 0.07884202152490616, + "learning_rate": 5.315293426175287e-06, + "loss": 0.0117, + "step": 88470 + }, + { + "epoch": 25.74338085539715, + "grad_norm": 0.08410536497831345, + "learning_rate": 5.312137171602987e-06, + "loss": 0.0118, + "step": 88480 + }, + { + "epoch": 25.74629036950829, + "grad_norm": 0.08902568370103836, + "learning_rate": 5.308981652759946e-06, + "loss": 0.0132, + "step": 88490 + }, + { + "epoch": 25.749199883619436, + "grad_norm": 0.14117811620235443, + "learning_rate": 5.305826869885792e-06, + "loss": 0.0102, + "step": 88500 + }, + { + "epoch": 25.752109397730578, + "grad_norm": 0.09212297946214676, + "learning_rate": 5.302672823220107e-06, + "loss": 0.0112, + "step": 88510 + }, + { + "epoch": 25.755018911841724, + "grad_norm": 0.10552917420864105, + "learning_rate": 5.299519513002425e-06, + "loss": 0.0118, + "step": 88520 + }, + { + "epoch": 25.757928425952866, + "grad_norm": 0.0877046287059784, + "learning_rate": 5.296366939472222e-06, + "loss": 0.0121, + "step": 88530 + }, + { + "epoch": 25.760837940064008, + "grad_norm": 0.11854823678731918, + "learning_rate": 5.293215102868914e-06, + "loss": 0.0134, + "step": 88540 + }, + { + "epoch": 25.763747454175153, + "grad_norm": 0.0997321754693985, + "learning_rate": 5.290064003431852e-06, + "loss": 0.0121, + "step": 88550 + }, + { + "epoch": 25.766656968286295, + "grad_norm": 0.10242005437612534, + "learning_rate": 5.286913641400351e-06, + "loss": 0.0115, + "step": 88560 + }, + { + "epoch": 25.76956648239744, + "grad_norm": 0.07567281275987625, + "learning_rate": 5.2837640170136605e-06, + "loss": 0.0117, + "step": 88570 + }, + { + "epoch": 25.772475996508582, + "grad_norm": 0.10252887010574341, + "learning_rate": 5.28061513051097e-06, + "loss": 0.0125, + "step": 88580 + }, + { + "epoch": 25.775385510619728, + "grad_norm": 0.07606921344995499, + "learning_rate": 5.277466982131414e-06, + "loss": 0.0102, + "step": 88590 + }, + { + "epoch": 25.77829502473087, + "grad_norm": 0.22677850723266602, + "learning_rate": 5.274319572114085e-06, + "loss": 0.0131, + "step": 88600 + }, + { + "epoch": 25.78120453884201, + "grad_norm": 0.10978946834802628, + "learning_rate": 5.271172900697997e-06, + "loss": 0.0117, + "step": 88610 + }, + { + "epoch": 25.784114052953157, + "grad_norm": 0.08862143009901047, + "learning_rate": 5.268026968122125e-06, + "loss": 0.0146, + "step": 88620 + }, + { + "epoch": 25.7870235670643, + "grad_norm": 0.10283005982637405, + "learning_rate": 5.264881774625375e-06, + "loss": 0.0115, + "step": 88630 + }, + { + "epoch": 25.789933081175445, + "grad_norm": 0.17887049913406372, + "learning_rate": 5.261737320446613e-06, + "loss": 0.0113, + "step": 88640 + }, + { + "epoch": 25.792842595286587, + "grad_norm": 0.07957135885953903, + "learning_rate": 5.2585936058246284e-06, + "loss": 0.0105, + "step": 88650 + }, + { + "epoch": 25.795752109397732, + "grad_norm": 0.10643641650676727, + "learning_rate": 5.255450630998179e-06, + "loss": 0.0109, + "step": 88660 + }, + { + "epoch": 25.798661623508874, + "grad_norm": 0.08920148015022278, + "learning_rate": 5.252308396205938e-06, + "loss": 0.0114, + "step": 88670 + }, + { + "epoch": 25.801571137620016, + "grad_norm": 0.09693536162376404, + "learning_rate": 5.249166901686548e-06, + "loss": 0.0111, + "step": 88680 + }, + { + "epoch": 25.80448065173116, + "grad_norm": 0.1283738762140274, + "learning_rate": 5.2460261476785855e-06, + "loss": 0.0123, + "step": 88690 + }, + { + "epoch": 25.807390165842303, + "grad_norm": 0.08173682540655136, + "learning_rate": 5.242886134420562e-06, + "loss": 0.0118, + "step": 88700 + }, + { + "epoch": 25.81029967995345, + "grad_norm": 0.10182774811983109, + "learning_rate": 5.2397468621509455e-06, + "loss": 0.0114, + "step": 88710 + }, + { + "epoch": 25.81320919406459, + "grad_norm": 0.10274915397167206, + "learning_rate": 5.236608331108147e-06, + "loss": 0.0128, + "step": 88720 + }, + { + "epoch": 25.816118708175736, + "grad_norm": 0.09294171631336212, + "learning_rate": 5.2334705415305125e-06, + "loss": 0.0131, + "step": 88730 + }, + { + "epoch": 25.819028222286878, + "grad_norm": 0.11129912734031677, + "learning_rate": 5.230333493656333e-06, + "loss": 0.0131, + "step": 88740 + }, + { + "epoch": 25.82193773639802, + "grad_norm": 0.12304433435201645, + "learning_rate": 5.22719718772385e-06, + "loss": 0.0127, + "step": 88750 + }, + { + "epoch": 25.824847250509166, + "grad_norm": 0.09368003904819489, + "learning_rate": 5.22406162397125e-06, + "loss": 0.013, + "step": 88760 + }, + { + "epoch": 25.827756764620307, + "grad_norm": 0.2238706648349762, + "learning_rate": 5.220926802636656e-06, + "loss": 0.0119, + "step": 88770 + }, + { + "epoch": 25.830666278731453, + "grad_norm": 0.11115951091051102, + "learning_rate": 5.217792723958123e-06, + "loss": 0.0121, + "step": 88780 + }, + { + "epoch": 25.833575792842595, + "grad_norm": 0.10679587721824646, + "learning_rate": 5.2146593881736865e-06, + "loss": 0.0104, + "step": 88790 + }, + { + "epoch": 25.83648530695374, + "grad_norm": 0.10187658667564392, + "learning_rate": 5.211526795521293e-06, + "loss": 0.0123, + "step": 88800 + }, + { + "epoch": 25.839394821064882, + "grad_norm": 0.10133921355009079, + "learning_rate": 5.2083949462388346e-06, + "loss": 0.0107, + "step": 88810 + }, + { + "epoch": 25.842304335176024, + "grad_norm": 0.12432372570037842, + "learning_rate": 5.205263840564161e-06, + "loss": 0.011, + "step": 88820 + }, + { + "epoch": 25.84521384928717, + "grad_norm": 0.11117817461490631, + "learning_rate": 5.202133478735065e-06, + "loss": 0.009, + "step": 88830 + }, + { + "epoch": 25.84812336339831, + "grad_norm": 0.09145641326904297, + "learning_rate": 5.199003860989271e-06, + "loss": 0.0133, + "step": 88840 + }, + { + "epoch": 25.851032877509457, + "grad_norm": 0.23598967492580414, + "learning_rate": 5.1958749875644455e-06, + "loss": 0.0133, + "step": 88850 + }, + { + "epoch": 25.8539423916206, + "grad_norm": 0.10682088136672974, + "learning_rate": 5.192746858698216e-06, + "loss": 0.0129, + "step": 88860 + }, + { + "epoch": 25.856851905731745, + "grad_norm": 0.09650710970163345, + "learning_rate": 5.189619474628144e-06, + "loss": 0.0136, + "step": 88870 + }, + { + "epoch": 25.859761419842886, + "grad_norm": 0.10322786122560501, + "learning_rate": 5.186492835591725e-06, + "loss": 0.0136, + "step": 88880 + }, + { + "epoch": 25.86267093395403, + "grad_norm": 0.114606574177742, + "learning_rate": 5.183366941826413e-06, + "loss": 0.0138, + "step": 88890 + }, + { + "epoch": 25.865580448065174, + "grad_norm": 0.12629081308841705, + "learning_rate": 5.1802417935696e-06, + "loss": 0.0103, + "step": 88900 + }, + { + "epoch": 25.868489962176316, + "grad_norm": 0.07543253898620605, + "learning_rate": 5.177117391058615e-06, + "loss": 0.011, + "step": 88910 + }, + { + "epoch": 25.87139947628746, + "grad_norm": 0.10297894477844238, + "learning_rate": 5.1739937345307435e-06, + "loss": 0.0132, + "step": 88920 + }, + { + "epoch": 25.874308990398603, + "grad_norm": 0.11241215467453003, + "learning_rate": 5.1708708242231965e-06, + "loss": 0.0134, + "step": 88930 + }, + { + "epoch": 25.87721850450975, + "grad_norm": 0.14580558240413666, + "learning_rate": 5.16774866037315e-06, + "loss": 0.0124, + "step": 88940 + }, + { + "epoch": 25.88012801862089, + "grad_norm": 0.10049839317798615, + "learning_rate": 5.164627243217698e-06, + "loss": 0.0129, + "step": 88950 + }, + { + "epoch": 25.883037532732033, + "grad_norm": 0.14725244045257568, + "learning_rate": 5.161506572993907e-06, + "loss": 0.0115, + "step": 88960 + }, + { + "epoch": 25.885947046843178, + "grad_norm": 0.08832670748233795, + "learning_rate": 5.158386649938756e-06, + "loss": 0.0123, + "step": 88970 + }, + { + "epoch": 25.88885656095432, + "grad_norm": 0.08706319332122803, + "learning_rate": 5.1552674742891906e-06, + "loss": 0.0097, + "step": 88980 + }, + { + "epoch": 25.891766075065465, + "grad_norm": 0.08671584725379944, + "learning_rate": 5.152149046282097e-06, + "loss": 0.0095, + "step": 88990 + }, + { + "epoch": 25.894675589176607, + "grad_norm": 0.11073006689548492, + "learning_rate": 5.1490313661542886e-06, + "loss": 0.0114, + "step": 89000 + }, + { + "epoch": 25.89758510328775, + "grad_norm": 0.14182555675506592, + "learning_rate": 5.14591443414254e-06, + "loss": 0.0139, + "step": 89010 + }, + { + "epoch": 25.900494617398895, + "grad_norm": 0.17721961438655853, + "learning_rate": 5.142798250483555e-06, + "loss": 0.0117, + "step": 89020 + }, + { + "epoch": 25.903404131510037, + "grad_norm": 0.08609306067228317, + "learning_rate": 5.139682815413996e-06, + "loss": 0.0131, + "step": 89030 + }, + { + "epoch": 25.906313645621182, + "grad_norm": 0.09551149606704712, + "learning_rate": 5.136568129170449e-06, + "loss": 0.0119, + "step": 89040 + }, + { + "epoch": 25.909223159732324, + "grad_norm": 0.10553261637687683, + "learning_rate": 5.1334541919894594e-06, + "loss": 0.014, + "step": 89050 + }, + { + "epoch": 25.91213267384347, + "grad_norm": 0.12800158560276031, + "learning_rate": 5.130341004107518e-06, + "loss": 0.0113, + "step": 89060 + }, + { + "epoch": 25.91504218795461, + "grad_norm": 0.0991319790482521, + "learning_rate": 5.12722856576104e-06, + "loss": 0.0147, + "step": 89070 + }, + { + "epoch": 25.917951702065753, + "grad_norm": 0.09480152279138565, + "learning_rate": 5.1241168771863895e-06, + "loss": 0.0115, + "step": 89080 + }, + { + "epoch": 25.9208612161769, + "grad_norm": 0.11939186602830887, + "learning_rate": 5.121005938619897e-06, + "loss": 0.0142, + "step": 89090 + }, + { + "epoch": 25.92377073028804, + "grad_norm": 0.08479515463113785, + "learning_rate": 5.117895750297807e-06, + "loss": 0.0133, + "step": 89100 + }, + { + "epoch": 25.926680244399186, + "grad_norm": 0.1367853432893753, + "learning_rate": 5.1147863124563145e-06, + "loss": 0.0114, + "step": 89110 + }, + { + "epoch": 25.92958975851033, + "grad_norm": 0.16290047764778137, + "learning_rate": 5.111677625331565e-06, + "loss": 0.0132, + "step": 89120 + }, + { + "epoch": 25.932499272621474, + "grad_norm": 0.0941636934876442, + "learning_rate": 5.108569689159646e-06, + "loss": 0.0135, + "step": 89130 + }, + { + "epoch": 25.935408786732616, + "grad_norm": 0.05854180455207825, + "learning_rate": 5.105462504176583e-06, + "loss": 0.0106, + "step": 89140 + }, + { + "epoch": 25.938318300843758, + "grad_norm": 0.07304222136735916, + "learning_rate": 5.102356070618338e-06, + "loss": 0.0097, + "step": 89150 + }, + { + "epoch": 25.941227814954903, + "grad_norm": 0.1427576243877411, + "learning_rate": 5.099250388720832e-06, + "loss": 0.0099, + "step": 89160 + }, + { + "epoch": 25.944137329066045, + "grad_norm": 0.10147321224212646, + "learning_rate": 5.096145458719924e-06, + "loss": 0.014, + "step": 89170 + }, + { + "epoch": 25.94704684317719, + "grad_norm": 0.09261626750230789, + "learning_rate": 5.0930412808514045e-06, + "loss": 0.0119, + "step": 89180 + }, + { + "epoch": 25.949956357288332, + "grad_norm": 0.07010843604803085, + "learning_rate": 5.089937855351019e-06, + "loss": 0.0124, + "step": 89190 + }, + { + "epoch": 25.952865871399478, + "grad_norm": 0.07714640349149704, + "learning_rate": 5.086835182454458e-06, + "loss": 0.011, + "step": 89200 + }, + { + "epoch": 25.95577538551062, + "grad_norm": 0.10557691007852554, + "learning_rate": 5.083733262397342e-06, + "loss": 0.0113, + "step": 89210 + }, + { + "epoch": 25.958684899621762, + "grad_norm": 0.10042458027601242, + "learning_rate": 5.0806320954152394e-06, + "loss": 0.0122, + "step": 89220 + }, + { + "epoch": 25.961594413732907, + "grad_norm": 0.12709765136241913, + "learning_rate": 5.077531681743667e-06, + "loss": 0.0122, + "step": 89230 + }, + { + "epoch": 25.96450392784405, + "grad_norm": 0.12762664258480072, + "learning_rate": 5.074432021618086e-06, + "loss": 0.0124, + "step": 89240 + }, + { + "epoch": 25.967413441955195, + "grad_norm": 0.0882437527179718, + "learning_rate": 5.071333115273885e-06, + "loss": 0.0123, + "step": 89250 + }, + { + "epoch": 25.970322956066337, + "grad_norm": 0.13078820705413818, + "learning_rate": 5.068234962946416e-06, + "loss": 0.0117, + "step": 89260 + }, + { + "epoch": 25.97323247017748, + "grad_norm": 0.07908745855093002, + "learning_rate": 5.065137564870952e-06, + "loss": 0.0117, + "step": 89270 + }, + { + "epoch": 25.976141984288624, + "grad_norm": 0.063503198325634, + "learning_rate": 5.062040921282725e-06, + "loss": 0.0103, + "step": 89280 + }, + { + "epoch": 25.979051498399766, + "grad_norm": 0.10239052772521973, + "learning_rate": 5.058945032416909e-06, + "loss": 0.0115, + "step": 89290 + }, + { + "epoch": 25.98196101251091, + "grad_norm": 0.10790276527404785, + "learning_rate": 5.05584989850861e-06, + "loss": 0.0107, + "step": 89300 + }, + { + "epoch": 25.984870526622053, + "grad_norm": 0.1208251416683197, + "learning_rate": 5.052755519792889e-06, + "loss": 0.0092, + "step": 89310 + }, + { + "epoch": 25.9877800407332, + "grad_norm": 0.10080737620592117, + "learning_rate": 5.049661896504736e-06, + "loss": 0.012, + "step": 89320 + }, + { + "epoch": 25.99068955484434, + "grad_norm": 0.10869285464286804, + "learning_rate": 5.0465690288791e-06, + "loss": 0.0102, + "step": 89330 + }, + { + "epoch": 25.993599068955483, + "grad_norm": 0.11024977266788483, + "learning_rate": 5.043476917150854e-06, + "loss": 0.0103, + "step": 89340 + }, + { + "epoch": 25.99650858306663, + "grad_norm": 0.07004570215940475, + "learning_rate": 5.040385561554829e-06, + "loss": 0.0104, + "step": 89350 + }, + { + "epoch": 25.99941809717777, + "grad_norm": 0.11394871771335602, + "learning_rate": 5.037294962325799e-06, + "loss": 0.0102, + "step": 89360 + }, + { + "epoch": 26.002327611288916, + "grad_norm": 0.06968764215707779, + "learning_rate": 5.034205119698466e-06, + "loss": 0.0112, + "step": 89370 + }, + { + "epoch": 26.005237125400058, + "grad_norm": 0.0978279635310173, + "learning_rate": 5.031116033907483e-06, + "loss": 0.0118, + "step": 89380 + }, + { + "epoch": 26.008146639511203, + "grad_norm": 0.08756644278764725, + "learning_rate": 5.028027705187448e-06, + "loss": 0.0112, + "step": 89390 + }, + { + "epoch": 26.011056153622345, + "grad_norm": 0.10527973622083664, + "learning_rate": 5.024940133772904e-06, + "loss": 0.0122, + "step": 89400 + }, + { + "epoch": 26.013965667733487, + "grad_norm": 0.14659753441810608, + "learning_rate": 5.0218533198983215e-06, + "loss": 0.0126, + "step": 89410 + }, + { + "epoch": 26.016875181844632, + "grad_norm": 0.12372612208127975, + "learning_rate": 5.018767263798132e-06, + "loss": 0.0124, + "step": 89420 + }, + { + "epoch": 26.019784695955774, + "grad_norm": 0.09608685970306396, + "learning_rate": 5.0156819657067e-06, + "loss": 0.0119, + "step": 89430 + }, + { + "epoch": 26.02269421006692, + "grad_norm": 0.12133630365133286, + "learning_rate": 5.012597425858334e-06, + "loss": 0.0133, + "step": 89440 + }, + { + "epoch": 26.025603724178062, + "grad_norm": 0.09444485604763031, + "learning_rate": 5.009513644487276e-06, + "loss": 0.011, + "step": 89450 + }, + { + "epoch": 26.028513238289207, + "grad_norm": 0.10371816158294678, + "learning_rate": 5.006430621827726e-06, + "loss": 0.0108, + "step": 89460 + }, + { + "epoch": 26.03142275240035, + "grad_norm": 0.07457626610994339, + "learning_rate": 5.003348358113824e-06, + "loss": 0.0097, + "step": 89470 + }, + { + "epoch": 26.03433226651149, + "grad_norm": 0.12135842442512512, + "learning_rate": 5.000266853579641e-06, + "loss": 0.0107, + "step": 89480 + }, + { + "epoch": 26.037241780622637, + "grad_norm": 0.136338010430336, + "learning_rate": 4.997186108459188e-06, + "loss": 0.0125, + "step": 89490 + }, + { + "epoch": 26.04015129473378, + "grad_norm": 0.07690443098545074, + "learning_rate": 4.994106122986448e-06, + "loss": 0.011, + "step": 89500 + }, + { + "epoch": 26.043060808844924, + "grad_norm": 0.08546865731477737, + "learning_rate": 4.991026897395314e-06, + "loss": 0.0097, + "step": 89510 + }, + { + "epoch": 26.045970322956066, + "grad_norm": 0.10333948582410812, + "learning_rate": 4.98794843191963e-06, + "loss": 0.01, + "step": 89520 + }, + { + "epoch": 26.04887983706721, + "grad_norm": 0.0968593955039978, + "learning_rate": 4.984870726793188e-06, + "loss": 0.0087, + "step": 89530 + }, + { + "epoch": 26.051789351178353, + "grad_norm": 0.09621843695640564, + "learning_rate": 4.981793782249725e-06, + "loss": 0.0156, + "step": 89540 + }, + { + "epoch": 26.054698865289495, + "grad_norm": 0.10237745940685272, + "learning_rate": 4.978717598522905e-06, + "loss": 0.011, + "step": 89550 + }, + { + "epoch": 26.05760837940064, + "grad_norm": 0.08376898616552353, + "learning_rate": 4.975642175846352e-06, + "loss": 0.0105, + "step": 89560 + }, + { + "epoch": 26.060517893511783, + "grad_norm": 0.08209507167339325, + "learning_rate": 4.97256751445362e-06, + "loss": 0.0115, + "step": 89570 + }, + { + "epoch": 26.063427407622928, + "grad_norm": 0.10036876052618027, + "learning_rate": 4.969493614578211e-06, + "loss": 0.0134, + "step": 89580 + }, + { + "epoch": 26.06633692173407, + "grad_norm": 0.12767066061496735, + "learning_rate": 4.966420476453563e-06, + "loss": 0.0103, + "step": 89590 + }, + { + "epoch": 26.069246435845216, + "grad_norm": 0.12127848714590073, + "learning_rate": 4.963348100313064e-06, + "loss": 0.013, + "step": 89600 + }, + { + "epoch": 26.072155949956358, + "grad_norm": 0.1078428328037262, + "learning_rate": 4.9602764863900445e-06, + "loss": 0.0114, + "step": 89610 + }, + { + "epoch": 26.0750654640675, + "grad_norm": 0.07218681275844574, + "learning_rate": 4.957205634917766e-06, + "loss": 0.0115, + "step": 89620 + }, + { + "epoch": 26.077974978178645, + "grad_norm": 0.08556217700242996, + "learning_rate": 4.954135546129448e-06, + "loss": 0.0124, + "step": 89630 + }, + { + "epoch": 26.080884492289787, + "grad_norm": 0.09663402289152145, + "learning_rate": 4.951066220258231e-06, + "loss": 0.011, + "step": 89640 + }, + { + "epoch": 26.083794006400932, + "grad_norm": 0.1496475636959076, + "learning_rate": 4.947997657537219e-06, + "loss": 0.0107, + "step": 89650 + }, + { + "epoch": 26.086703520512074, + "grad_norm": 0.08986788243055344, + "learning_rate": 4.944929858199452e-06, + "loss": 0.0116, + "step": 89660 + }, + { + "epoch": 26.089613034623216, + "grad_norm": 0.0975010022521019, + "learning_rate": 4.941862822477905e-06, + "loss": 0.0129, + "step": 89670 + }, + { + "epoch": 26.09252254873436, + "grad_norm": 0.1189657673239708, + "learning_rate": 4.938796550605494e-06, + "loss": 0.0105, + "step": 89680 + }, + { + "epoch": 26.095432062845504, + "grad_norm": 0.14361122250556946, + "learning_rate": 4.935731042815088e-06, + "loss": 0.0094, + "step": 89690 + }, + { + "epoch": 26.09834157695665, + "grad_norm": 0.11967582255601883, + "learning_rate": 4.932666299339494e-06, + "loss": 0.0107, + "step": 89700 + }, + { + "epoch": 26.10125109106779, + "grad_norm": 0.07836959511041641, + "learning_rate": 4.929602320411451e-06, + "loss": 0.011, + "step": 89710 + }, + { + "epoch": 26.104160605178937, + "grad_norm": 0.08054110407829285, + "learning_rate": 4.926539106263653e-06, + "loss": 0.0125, + "step": 89720 + }, + { + "epoch": 26.10707011929008, + "grad_norm": 0.17746509611606598, + "learning_rate": 4.923476657128736e-06, + "loss": 0.0117, + "step": 89730 + }, + { + "epoch": 26.10997963340122, + "grad_norm": 0.10785964876413345, + "learning_rate": 4.920414973239267e-06, + "loss": 0.0123, + "step": 89740 + }, + { + "epoch": 26.112889147512366, + "grad_norm": 0.14870581030845642, + "learning_rate": 4.917354054827757e-06, + "loss": 0.0106, + "step": 89750 + }, + { + "epoch": 26.115798661623508, + "grad_norm": 0.07345924526453018, + "learning_rate": 4.914293902126668e-06, + "loss": 0.0112, + "step": 89760 + }, + { + "epoch": 26.118708175734653, + "grad_norm": 0.08498101681470871, + "learning_rate": 4.9112345153684005e-06, + "loss": 0.0119, + "step": 89770 + }, + { + "epoch": 26.121617689845795, + "grad_norm": 0.08542249351739883, + "learning_rate": 4.908175894785293e-06, + "loss": 0.0134, + "step": 89780 + }, + { + "epoch": 26.12452720395694, + "grad_norm": 0.11035823076963425, + "learning_rate": 4.905118040609615e-06, + "loss": 0.0122, + "step": 89790 + }, + { + "epoch": 26.127436718068083, + "grad_norm": 0.10539872199296951, + "learning_rate": 4.9020609530736116e-06, + "loss": 0.011, + "step": 89800 + }, + { + "epoch": 26.130346232179225, + "grad_norm": 0.10010574758052826, + "learning_rate": 4.899004632409438e-06, + "loss": 0.0115, + "step": 89810 + }, + { + "epoch": 26.13325574629037, + "grad_norm": 0.0883035659790039, + "learning_rate": 4.895949078849196e-06, + "loss": 0.0108, + "step": 89820 + }, + { + "epoch": 26.136165260401512, + "grad_norm": 0.08121582120656967, + "learning_rate": 4.892894292624943e-06, + "loss": 0.0114, + "step": 89830 + }, + { + "epoch": 26.139074774512657, + "grad_norm": 0.12971028685569763, + "learning_rate": 4.8898402739686695e-06, + "loss": 0.0102, + "step": 89840 + }, + { + "epoch": 26.1419842886238, + "grad_norm": 0.1204080879688263, + "learning_rate": 4.886787023112308e-06, + "loss": 0.0104, + "step": 89850 + }, + { + "epoch": 26.144893802734945, + "grad_norm": 0.1499832719564438, + "learning_rate": 4.883734540287725e-06, + "loss": 0.0103, + "step": 89860 + }, + { + "epoch": 26.147803316846087, + "grad_norm": 0.12314704805612564, + "learning_rate": 4.880682825726742e-06, + "loss": 0.0133, + "step": 89870 + }, + { + "epoch": 26.15071283095723, + "grad_norm": 0.08866723626852036, + "learning_rate": 4.877631879661123e-06, + "loss": 0.0127, + "step": 89880 + }, + { + "epoch": 26.153622345068374, + "grad_norm": 0.17901040613651276, + "learning_rate": 4.874581702322553e-06, + "loss": 0.0101, + "step": 89890 + }, + { + "epoch": 26.156531859179516, + "grad_norm": 0.09928476065397263, + "learning_rate": 4.871532293942683e-06, + "loss": 0.011, + "step": 89900 + }, + { + "epoch": 26.15944137329066, + "grad_norm": 0.11077095568180084, + "learning_rate": 4.868483654753097e-06, + "loss": 0.011, + "step": 89910 + }, + { + "epoch": 26.162350887401804, + "grad_norm": 0.11131104826927185, + "learning_rate": 4.865435784985311e-06, + "loss": 0.0122, + "step": 89920 + }, + { + "epoch": 26.16526040151295, + "grad_norm": 0.10429958254098892, + "learning_rate": 4.8623886848708e-06, + "loss": 0.0131, + "step": 89930 + }, + { + "epoch": 26.16816991562409, + "grad_norm": 0.1763681173324585, + "learning_rate": 4.859342354640961e-06, + "loss": 0.0109, + "step": 89940 + }, + { + "epoch": 26.171079429735233, + "grad_norm": 0.09277047961950302, + "learning_rate": 4.856296794527153e-06, + "loss": 0.0116, + "step": 89950 + }, + { + "epoch": 26.17398894384638, + "grad_norm": 0.1342877596616745, + "learning_rate": 4.853252004760654e-06, + "loss": 0.0123, + "step": 89960 + }, + { + "epoch": 26.17689845795752, + "grad_norm": 0.09743522107601166, + "learning_rate": 4.85020798557271e-06, + "loss": 0.0116, + "step": 89970 + }, + { + "epoch": 26.179807972068666, + "grad_norm": 0.08068060129880905, + "learning_rate": 4.84716473719448e-06, + "loss": 0.0098, + "step": 89980 + }, + { + "epoch": 26.182717486179808, + "grad_norm": 0.1007978618144989, + "learning_rate": 4.844122259857086e-06, + "loss": 0.0107, + "step": 89990 + }, + { + "epoch": 26.18562700029095, + "grad_norm": 0.0839049369096756, + "learning_rate": 4.841080553791589e-06, + "loss": 0.0104, + "step": 90000 + }, + { + "epoch": 26.188536514402095, + "grad_norm": 0.08725588023662567, + "learning_rate": 4.838039619228974e-06, + "loss": 0.011, + "step": 90010 + }, + { + "epoch": 26.191446028513237, + "grad_norm": 0.1069418266415596, + "learning_rate": 4.834999456400188e-06, + "loss": 0.0109, + "step": 90020 + }, + { + "epoch": 26.194355542624383, + "grad_norm": 0.149014413356781, + "learning_rate": 4.831960065536115e-06, + "loss": 0.0113, + "step": 90030 + }, + { + "epoch": 26.197265056735525, + "grad_norm": 0.09270782023668289, + "learning_rate": 4.828921446867572e-06, + "loss": 0.011, + "step": 90040 + }, + { + "epoch": 26.20017457084667, + "grad_norm": 0.10007612407207489, + "learning_rate": 4.825883600625316e-06, + "loss": 0.0107, + "step": 90050 + }, + { + "epoch": 26.203084084957812, + "grad_norm": 0.11658672988414764, + "learning_rate": 4.822846527040058e-06, + "loss": 0.0115, + "step": 90060 + }, + { + "epoch": 26.205993599068954, + "grad_norm": 0.16269689798355103, + "learning_rate": 4.819810226342449e-06, + "loss": 0.0103, + "step": 90070 + }, + { + "epoch": 26.2089031131801, + "grad_norm": 0.11922622472047806, + "learning_rate": 4.816774698763067e-06, + "loss": 0.0109, + "step": 90080 + }, + { + "epoch": 26.21181262729124, + "grad_norm": 0.27552881836891174, + "learning_rate": 4.813739944532436e-06, + "loss": 0.0109, + "step": 90090 + }, + { + "epoch": 26.214722141402387, + "grad_norm": 0.08755198866128922, + "learning_rate": 4.810705963881042e-06, + "loss": 0.0118, + "step": 90100 + }, + { + "epoch": 26.21763165551353, + "grad_norm": 0.1502755731344223, + "learning_rate": 4.807672757039286e-06, + "loss": 0.0137, + "step": 90110 + }, + { + "epoch": 26.220541169624674, + "grad_norm": 0.19614984095096588, + "learning_rate": 4.8046403242375155e-06, + "loss": 0.0144, + "step": 90120 + }, + { + "epoch": 26.223450683735816, + "grad_norm": 0.11615870893001556, + "learning_rate": 4.801608665706029e-06, + "loss": 0.0105, + "step": 90130 + }, + { + "epoch": 26.226360197846958, + "grad_norm": 0.10994139313697815, + "learning_rate": 4.798577781675066e-06, + "loss": 0.0129, + "step": 90140 + }, + { + "epoch": 26.229269711958104, + "grad_norm": 0.10202840715646744, + "learning_rate": 4.795547672374798e-06, + "loss": 0.0116, + "step": 90150 + }, + { + "epoch": 26.232179226069245, + "grad_norm": 0.12712840735912323, + "learning_rate": 4.792518338035334e-06, + "loss": 0.015, + "step": 90160 + }, + { + "epoch": 26.23508874018039, + "grad_norm": 0.1922171711921692, + "learning_rate": 4.78948977888674e-06, + "loss": 0.0121, + "step": 90170 + }, + { + "epoch": 26.237998254291533, + "grad_norm": 0.13594913482666016, + "learning_rate": 4.786461995159017e-06, + "loss": 0.0099, + "step": 90180 + }, + { + "epoch": 26.24090776840268, + "grad_norm": 0.09519024938344955, + "learning_rate": 4.7834349870820984e-06, + "loss": 0.0114, + "step": 90190 + }, + { + "epoch": 26.24381728251382, + "grad_norm": 0.13181206583976746, + "learning_rate": 4.780408754885869e-06, + "loss": 0.011, + "step": 90200 + }, + { + "epoch": 26.246726796624962, + "grad_norm": 0.09033214300870895, + "learning_rate": 4.777383298800157e-06, + "loss": 0.0124, + "step": 90210 + }, + { + "epoch": 26.249636310736108, + "grad_norm": 0.1790747344493866, + "learning_rate": 4.774358619054718e-06, + "loss": 0.0121, + "step": 90220 + }, + { + "epoch": 26.25254582484725, + "grad_norm": 0.12913504242897034, + "learning_rate": 4.771334715879254e-06, + "loss": 0.0128, + "step": 90230 + }, + { + "epoch": 26.255455338958395, + "grad_norm": 0.10624168068170547, + "learning_rate": 4.768311589503417e-06, + "loss": 0.0122, + "step": 90240 + }, + { + "epoch": 26.258364853069537, + "grad_norm": 0.10724175721406937, + "learning_rate": 4.765289240156794e-06, + "loss": 0.0121, + "step": 90250 + }, + { + "epoch": 26.261274367180683, + "grad_norm": 0.08217649161815643, + "learning_rate": 4.762267668068906e-06, + "loss": 0.0119, + "step": 90260 + }, + { + "epoch": 26.264183881291824, + "grad_norm": 0.10530491173267365, + "learning_rate": 4.759246873469231e-06, + "loss": 0.0119, + "step": 90270 + }, + { + "epoch": 26.267093395402966, + "grad_norm": 0.09962466359138489, + "learning_rate": 4.756226856587168e-06, + "loss": 0.0115, + "step": 90280 + }, + { + "epoch": 26.270002909514112, + "grad_norm": 0.08870021253824234, + "learning_rate": 4.753207617652072e-06, + "loss": 0.0123, + "step": 90290 + }, + { + "epoch": 26.272912423625254, + "grad_norm": 0.10023543238639832, + "learning_rate": 4.7501891568932385e-06, + "loss": 0.0108, + "step": 90300 + }, + { + "epoch": 26.2758219377364, + "grad_norm": 0.14649999141693115, + "learning_rate": 4.747171474539891e-06, + "loss": 0.0106, + "step": 90310 + }, + { + "epoch": 26.27873145184754, + "grad_norm": 0.2742539644241333, + "learning_rate": 4.744154570821213e-06, + "loss": 0.0122, + "step": 90320 + }, + { + "epoch": 26.281640965958683, + "grad_norm": 0.1050809994339943, + "learning_rate": 4.741138445966307e-06, + "loss": 0.0123, + "step": 90330 + }, + { + "epoch": 26.28455048006983, + "grad_norm": 0.13148674368858337, + "learning_rate": 4.738123100204238e-06, + "loss": 0.013, + "step": 90340 + }, + { + "epoch": 26.28745999418097, + "grad_norm": 0.5291737914085388, + "learning_rate": 4.7351085337639925e-06, + "loss": 0.0119, + "step": 90350 + }, + { + "epoch": 26.290369508292116, + "grad_norm": 0.15438531339168549, + "learning_rate": 4.732094746874513e-06, + "loss": 0.0117, + "step": 90360 + }, + { + "epoch": 26.293279022403258, + "grad_norm": 0.08039068430662155, + "learning_rate": 4.729081739764678e-06, + "loss": 0.0104, + "step": 90370 + }, + { + "epoch": 26.296188536514403, + "grad_norm": 0.0983627662062645, + "learning_rate": 4.726069512663303e-06, + "loss": 0.0142, + "step": 90380 + }, + { + "epoch": 26.299098050625545, + "grad_norm": 0.08594179898500443, + "learning_rate": 4.723058065799139e-06, + "loss": 0.0114, + "step": 90390 + }, + { + "epoch": 26.302007564736687, + "grad_norm": 0.09103954583406448, + "learning_rate": 4.720047399400899e-06, + "loss": 0.0107, + "step": 90400 + }, + { + "epoch": 26.304917078847833, + "grad_norm": 0.11164502799510956, + "learning_rate": 4.717037513697221e-06, + "loss": 0.0124, + "step": 90410 + }, + { + "epoch": 26.307826592958975, + "grad_norm": 0.12137524783611298, + "learning_rate": 4.714028408916675e-06, + "loss": 0.0139, + "step": 90420 + }, + { + "epoch": 26.31073610707012, + "grad_norm": 0.10045269876718521, + "learning_rate": 4.711020085287791e-06, + "loss": 0.0114, + "step": 90430 + }, + { + "epoch": 26.313645621181262, + "grad_norm": 0.1035187840461731, + "learning_rate": 4.708012543039034e-06, + "loss": 0.0098, + "step": 90440 + }, + { + "epoch": 26.316555135292408, + "grad_norm": 0.0985151007771492, + "learning_rate": 4.705005782398803e-06, + "loss": 0.0115, + "step": 90450 + }, + { + "epoch": 26.31946464940355, + "grad_norm": 0.12011783570051193, + "learning_rate": 4.701999803595437e-06, + "loss": 0.0134, + "step": 90460 + }, + { + "epoch": 26.32237416351469, + "grad_norm": 0.10702847689390182, + "learning_rate": 4.698994606857225e-06, + "loss": 0.0142, + "step": 90470 + }, + { + "epoch": 26.325283677625837, + "grad_norm": 0.09349724650382996, + "learning_rate": 4.695990192412397e-06, + "loss": 0.0136, + "step": 90480 + }, + { + "epoch": 26.32819319173698, + "grad_norm": 0.11306893825531006, + "learning_rate": 4.692986560489108e-06, + "loss": 0.0097, + "step": 90490 + }, + { + "epoch": 26.331102705848124, + "grad_norm": 0.13912679255008698, + "learning_rate": 4.689983711315472e-06, + "loss": 0.0098, + "step": 90500 + }, + { + "epoch": 26.334012219959266, + "grad_norm": 0.1072721928358078, + "learning_rate": 4.6869816451195346e-06, + "loss": 0.0118, + "step": 90510 + }, + { + "epoch": 26.336921734070412, + "grad_norm": 0.07842212170362473, + "learning_rate": 4.683980362129283e-06, + "loss": 0.0105, + "step": 90520 + }, + { + "epoch": 26.339831248181554, + "grad_norm": 0.08809266239404678, + "learning_rate": 4.680979862572639e-06, + "loss": 0.0111, + "step": 90530 + }, + { + "epoch": 26.342740762292696, + "grad_norm": 0.14273081719875336, + "learning_rate": 4.677980146677472e-06, + "loss": 0.0134, + "step": 90540 + }, + { + "epoch": 26.34565027640384, + "grad_norm": 0.11116701364517212, + "learning_rate": 4.6749812146716005e-06, + "loss": 0.0139, + "step": 90550 + }, + { + "epoch": 26.348559790514983, + "grad_norm": 0.09358370304107666, + "learning_rate": 4.6719830667827615e-06, + "loss": 0.0138, + "step": 90560 + }, + { + "epoch": 26.35146930462613, + "grad_norm": 0.1163971871137619, + "learning_rate": 4.668985703238654e-06, + "loss": 0.011, + "step": 90570 + }, + { + "epoch": 26.35437881873727, + "grad_norm": 0.09496762603521347, + "learning_rate": 4.665989124266898e-06, + "loss": 0.0093, + "step": 90580 + }, + { + "epoch": 26.357288332848416, + "grad_norm": 0.1250516027212143, + "learning_rate": 4.662993330095076e-06, + "loss": 0.0122, + "step": 90590 + }, + { + "epoch": 26.360197846959558, + "grad_norm": 0.1159338429570198, + "learning_rate": 4.659998320950685e-06, + "loss": 0.0122, + "step": 90600 + }, + { + "epoch": 26.3631073610707, + "grad_norm": 0.12476833909749985, + "learning_rate": 4.657004097061186e-06, + "loss": 0.0122, + "step": 90610 + }, + { + "epoch": 26.366016875181845, + "grad_norm": 0.10276801139116287, + "learning_rate": 4.6540106586539714e-06, + "loss": 0.012, + "step": 90620 + }, + { + "epoch": 26.368926389292987, + "grad_norm": 0.11106456816196442, + "learning_rate": 4.6510180059563654e-06, + "loss": 0.0117, + "step": 90630 + }, + { + "epoch": 26.371835903404133, + "grad_norm": 0.09830178320407867, + "learning_rate": 4.648026139195649e-06, + "loss": 0.0126, + "step": 90640 + }, + { + "epoch": 26.374745417515275, + "grad_norm": 0.09898657351732254, + "learning_rate": 4.645035058599025e-06, + "loss": 0.0103, + "step": 90650 + }, + { + "epoch": 26.377654931626417, + "grad_norm": 0.09150796383619308, + "learning_rate": 4.6420447643936534e-06, + "loss": 0.0134, + "step": 90660 + }, + { + "epoch": 26.380564445737562, + "grad_norm": 0.11796854436397552, + "learning_rate": 4.639055256806629e-06, + "loss": 0.0113, + "step": 90670 + }, + { + "epoch": 26.383473959848704, + "grad_norm": 0.17374378442764282, + "learning_rate": 4.63606653606498e-06, + "loss": 0.0111, + "step": 90680 + }, + { + "epoch": 26.38638347395985, + "grad_norm": 0.12487521022558212, + "learning_rate": 4.6330786023956786e-06, + "loss": 0.013, + "step": 90690 + }, + { + "epoch": 26.38929298807099, + "grad_norm": 0.12524235248565674, + "learning_rate": 4.630091456025641e-06, + "loss": 0.0119, + "step": 90700 + }, + { + "epoch": 26.392202502182137, + "grad_norm": 0.09337666630744934, + "learning_rate": 4.627105097181728e-06, + "loss": 0.0096, + "step": 90710 + }, + { + "epoch": 26.39511201629328, + "grad_norm": 0.10022281110286713, + "learning_rate": 4.624119526090722e-06, + "loss": 0.0117, + "step": 90720 + }, + { + "epoch": 26.39802153040442, + "grad_norm": 0.11115587502717972, + "learning_rate": 4.621134742979361e-06, + "loss": 0.0119, + "step": 90730 + }, + { + "epoch": 26.400931044515566, + "grad_norm": 0.1378472000360489, + "learning_rate": 4.618150748074329e-06, + "loss": 0.0122, + "step": 90740 + }, + { + "epoch": 26.403840558626708, + "grad_norm": 0.10111185908317566, + "learning_rate": 4.615167541602233e-06, + "loss": 0.0114, + "step": 90750 + }, + { + "epoch": 26.406750072737854, + "grad_norm": 0.10256028920412064, + "learning_rate": 4.6121851237896244e-06, + "loss": 0.0114, + "step": 90760 + }, + { + "epoch": 26.409659586848996, + "grad_norm": 0.08003424108028412, + "learning_rate": 4.609203494863e-06, + "loss": 0.0096, + "step": 90770 + }, + { + "epoch": 26.41256910096014, + "grad_norm": 0.1134411096572876, + "learning_rate": 4.606222655048804e-06, + "loss": 0.0114, + "step": 90780 + }, + { + "epoch": 26.415478615071283, + "grad_norm": 0.08122224360704422, + "learning_rate": 4.603242604573403e-06, + "loss": 0.0111, + "step": 90790 + }, + { + "epoch": 26.418388129182425, + "grad_norm": 0.091563381254673, + "learning_rate": 4.600263343663104e-06, + "loss": 0.0103, + "step": 90800 + }, + { + "epoch": 26.42129764329357, + "grad_norm": 0.12655572593212128, + "learning_rate": 4.597284872544181e-06, + "loss": 0.0106, + "step": 90810 + }, + { + "epoch": 26.424207157404712, + "grad_norm": 0.0899205282330513, + "learning_rate": 4.59430719144282e-06, + "loss": 0.009, + "step": 90820 + }, + { + "epoch": 26.427116671515858, + "grad_norm": 0.08645036071538925, + "learning_rate": 4.591330300585152e-06, + "loss": 0.0133, + "step": 90830 + }, + { + "epoch": 26.430026185627, + "grad_norm": 0.0885079950094223, + "learning_rate": 4.588354200197257e-06, + "loss": 0.0105, + "step": 90840 + }, + { + "epoch": 26.432935699738145, + "grad_norm": 0.3646257817745209, + "learning_rate": 4.585378890505154e-06, + "loss": 0.0097, + "step": 90850 + }, + { + "epoch": 26.435845213849287, + "grad_norm": 0.10318149626255035, + "learning_rate": 4.582404371734787e-06, + "loss": 0.0122, + "step": 90860 + }, + { + "epoch": 26.43875472796043, + "grad_norm": 0.08863361924886703, + "learning_rate": 4.5794306441120645e-06, + "loss": 0.011, + "step": 90870 + }, + { + "epoch": 26.441664242071575, + "grad_norm": 0.09441103786230087, + "learning_rate": 4.57645770786281e-06, + "loss": 0.0113, + "step": 90880 + }, + { + "epoch": 26.444573756182717, + "grad_norm": 0.1265801042318344, + "learning_rate": 4.573485563212807e-06, + "loss": 0.0115, + "step": 90890 + }, + { + "epoch": 26.447483270293862, + "grad_norm": 0.10067295283079147, + "learning_rate": 4.5705142103877626e-06, + "loss": 0.0109, + "step": 90900 + }, + { + "epoch": 26.450392784405004, + "grad_norm": 0.18609189987182617, + "learning_rate": 4.567543649613336e-06, + "loss": 0.0117, + "step": 90910 + }, + { + "epoch": 26.45330229851615, + "grad_norm": 0.10647151619195938, + "learning_rate": 4.564573881115124e-06, + "loss": 0.0108, + "step": 90920 + }, + { + "epoch": 26.45621181262729, + "grad_norm": 0.08267899602651596, + "learning_rate": 4.561604905118654e-06, + "loss": 0.0116, + "step": 90930 + }, + { + "epoch": 26.459121326738433, + "grad_norm": 0.09946750104427338, + "learning_rate": 4.558636721849409e-06, + "loss": 0.0132, + "step": 90940 + }, + { + "epoch": 26.46203084084958, + "grad_norm": 0.10589633136987686, + "learning_rate": 4.555669331532793e-06, + "loss": 0.0142, + "step": 90950 + }, + { + "epoch": 26.46494035496072, + "grad_norm": 0.10588102787733078, + "learning_rate": 4.5527027343941655e-06, + "loss": 0.0135, + "step": 90960 + }, + { + "epoch": 26.467849869071866, + "grad_norm": 0.22374524176120758, + "learning_rate": 4.549736930658823e-06, + "loss": 0.0112, + "step": 90970 + }, + { + "epoch": 26.470759383183008, + "grad_norm": 0.12799575924873352, + "learning_rate": 4.5467719205519945e-06, + "loss": 0.0129, + "step": 90980 + }, + { + "epoch": 26.473668897294154, + "grad_norm": 0.1023198589682579, + "learning_rate": 4.543807704298849e-06, + "loss": 0.013, + "step": 90990 + }, + { + "epoch": 26.476578411405296, + "grad_norm": 0.1371038556098938, + "learning_rate": 4.540844282124504e-06, + "loss": 0.0115, + "step": 91000 + }, + { + "epoch": 26.479487925516437, + "grad_norm": 0.0804385393857956, + "learning_rate": 4.537881654254015e-06, + "loss": 0.0125, + "step": 91010 + }, + { + "epoch": 26.482397439627583, + "grad_norm": 0.0879087895154953, + "learning_rate": 4.534919820912366e-06, + "loss": 0.0102, + "step": 91020 + }, + { + "epoch": 26.485306953738725, + "grad_norm": 0.12771983444690704, + "learning_rate": 4.531958782324495e-06, + "loss": 0.0134, + "step": 91030 + }, + { + "epoch": 26.48821646784987, + "grad_norm": 0.09454658627510071, + "learning_rate": 4.528998538715274e-06, + "loss": 0.0108, + "step": 91040 + }, + { + "epoch": 26.491125981961012, + "grad_norm": 0.07814504206180573, + "learning_rate": 4.526039090309512e-06, + "loss": 0.0109, + "step": 91050 + }, + { + "epoch": 26.494035496072154, + "grad_norm": 0.1336139738559723, + "learning_rate": 4.523080437331953e-06, + "loss": 0.0133, + "step": 91060 + }, + { + "epoch": 26.4969450101833, + "grad_norm": 0.16186927258968353, + "learning_rate": 4.520122580007295e-06, + "loss": 0.0134, + "step": 91070 + }, + { + "epoch": 26.49985452429444, + "grad_norm": 0.13468362390995026, + "learning_rate": 4.51716551856017e-06, + "loss": 0.0122, + "step": 91080 + }, + { + "epoch": 26.502764038405587, + "grad_norm": 0.08516331762075424, + "learning_rate": 4.514209253215143e-06, + "loss": 0.0102, + "step": 91090 + }, + { + "epoch": 26.50567355251673, + "grad_norm": 0.1115460991859436, + "learning_rate": 4.511253784196714e-06, + "loss": 0.0126, + "step": 91100 + }, + { + "epoch": 26.508583066627875, + "grad_norm": 0.10566014796495438, + "learning_rate": 4.50829911172935e-06, + "loss": 0.0092, + "step": 91110 + }, + { + "epoch": 26.511492580739016, + "grad_norm": 0.09678252786397934, + "learning_rate": 4.505345236037428e-06, + "loss": 0.0092, + "step": 91120 + }, + { + "epoch": 26.51440209485016, + "grad_norm": 0.12016180902719498, + "learning_rate": 4.502392157345272e-06, + "loss": 0.0107, + "step": 91130 + }, + { + "epoch": 26.517311608961304, + "grad_norm": 0.12916220724582672, + "learning_rate": 4.4994398758771555e-06, + "loss": 0.0106, + "step": 91140 + }, + { + "epoch": 26.520221123072446, + "grad_norm": 0.09976735711097717, + "learning_rate": 4.496488391857285e-06, + "loss": 0.0134, + "step": 91150 + }, + { + "epoch": 26.52313063718359, + "grad_norm": 0.11443876475095749, + "learning_rate": 4.4935377055098055e-06, + "loss": 0.0111, + "step": 91160 + }, + { + "epoch": 26.526040151294733, + "grad_norm": 0.13520805537700653, + "learning_rate": 4.490587817058795e-06, + "loss": 0.0104, + "step": 91170 + }, + { + "epoch": 26.52894966540588, + "grad_norm": 0.1370030790567398, + "learning_rate": 4.487638726728285e-06, + "loss": 0.0131, + "step": 91180 + }, + { + "epoch": 26.53185917951702, + "grad_norm": 0.09787514805793762, + "learning_rate": 4.484690434742242e-06, + "loss": 0.0104, + "step": 91190 + }, + { + "epoch": 26.534768693628163, + "grad_norm": 0.08659069240093231, + "learning_rate": 4.481742941324561e-06, + "loss": 0.0098, + "step": 91200 + }, + { + "epoch": 26.537678207739308, + "grad_norm": 0.1023460105061531, + "learning_rate": 4.478796246699088e-06, + "loss": 0.0113, + "step": 91210 + }, + { + "epoch": 26.54058772185045, + "grad_norm": 0.07952608168125153, + "learning_rate": 4.475850351089613e-06, + "loss": 0.0092, + "step": 91220 + }, + { + "epoch": 26.543497235961595, + "grad_norm": 0.2599896192550659, + "learning_rate": 4.472905254719844e-06, + "loss": 0.0107, + "step": 91230 + }, + { + "epoch": 26.546406750072737, + "grad_norm": 0.11218587309122086, + "learning_rate": 4.469960957813453e-06, + "loss": 0.0117, + "step": 91240 + }, + { + "epoch": 26.549316264183883, + "grad_norm": 0.14861713349819183, + "learning_rate": 4.467017460594032e-06, + "loss": 0.0117, + "step": 91250 + }, + { + "epoch": 26.552225778295025, + "grad_norm": 0.17651528120040894, + "learning_rate": 4.464074763285126e-06, + "loss": 0.0102, + "step": 91260 + }, + { + "epoch": 26.555135292406167, + "grad_norm": 0.09837374836206436, + "learning_rate": 4.461132866110207e-06, + "loss": 0.0094, + "step": 91270 + }, + { + "epoch": 26.558044806517312, + "grad_norm": 0.09863891452550888, + "learning_rate": 4.458191769292701e-06, + "loss": 0.0119, + "step": 91280 + }, + { + "epoch": 26.560954320628454, + "grad_norm": 0.12719625234603882, + "learning_rate": 4.455251473055955e-06, + "loss": 0.011, + "step": 91290 + }, + { + "epoch": 26.5638638347396, + "grad_norm": 0.1400635540485382, + "learning_rate": 4.452311977623271e-06, + "loss": 0.0129, + "step": 91300 + }, + { + "epoch": 26.56677334885074, + "grad_norm": 0.1048617735505104, + "learning_rate": 4.449373283217887e-06, + "loss": 0.0109, + "step": 91310 + }, + { + "epoch": 26.569682862961887, + "grad_norm": 0.10718239098787308, + "learning_rate": 4.446435390062971e-06, + "loss": 0.0103, + "step": 91320 + }, + { + "epoch": 26.57259237707303, + "grad_norm": 0.11520854383707047, + "learning_rate": 4.4434982983816394e-06, + "loss": 0.0125, + "step": 91330 + }, + { + "epoch": 26.57550189118417, + "grad_norm": 0.11858239024877548, + "learning_rate": 4.44056200839695e-06, + "loss": 0.0101, + "step": 91340 + }, + { + "epoch": 26.578411405295316, + "grad_norm": 0.09878456592559814, + "learning_rate": 4.437626520331889e-06, + "loss": 0.0111, + "step": 91350 + }, + { + "epoch": 26.58132091940646, + "grad_norm": 0.12802881002426147, + "learning_rate": 4.434691834409384e-06, + "loss": 0.0115, + "step": 91360 + }, + { + "epoch": 26.584230433517604, + "grad_norm": 0.07962270826101303, + "learning_rate": 4.431757950852308e-06, + "loss": 0.0099, + "step": 91370 + }, + { + "epoch": 26.587139947628746, + "grad_norm": 0.08883122354745865, + "learning_rate": 4.428824869883478e-06, + "loss": 0.0129, + "step": 91380 + }, + { + "epoch": 26.590049461739888, + "grad_norm": 0.08238878101110458, + "learning_rate": 4.425892591725635e-06, + "loss": 0.0111, + "step": 91390 + }, + { + "epoch": 26.592958975851033, + "grad_norm": 0.11293130367994308, + "learning_rate": 4.422961116601457e-06, + "loss": 0.0113, + "step": 91400 + }, + { + "epoch": 26.595868489962175, + "grad_norm": 0.08979939669370651, + "learning_rate": 4.420030444733587e-06, + "loss": 0.0124, + "step": 91410 + }, + { + "epoch": 26.59877800407332, + "grad_norm": 0.12223154306411743, + "learning_rate": 4.417100576344586e-06, + "loss": 0.011, + "step": 91420 + }, + { + "epoch": 26.601687518184463, + "grad_norm": 0.09921611845493317, + "learning_rate": 4.414171511656948e-06, + "loss": 0.0114, + "step": 91430 + }, + { + "epoch": 26.604597032295608, + "grad_norm": 0.1262313723564148, + "learning_rate": 4.411243250893125e-06, + "loss": 0.0154, + "step": 91440 + }, + { + "epoch": 26.60750654640675, + "grad_norm": 0.09721556305885315, + "learning_rate": 4.408315794275502e-06, + "loss": 0.0106, + "step": 91450 + }, + { + "epoch": 26.610416060517892, + "grad_norm": 0.09846573323011398, + "learning_rate": 4.405389142026395e-06, + "loss": 0.0112, + "step": 91460 + }, + { + "epoch": 26.613325574629037, + "grad_norm": 0.11640197038650513, + "learning_rate": 4.402463294368056e-06, + "loss": 0.0124, + "step": 91470 + }, + { + "epoch": 26.61623508874018, + "grad_norm": 0.14038334786891937, + "learning_rate": 4.399538251522701e-06, + "loss": 0.01, + "step": 91480 + }, + { + "epoch": 26.619144602851325, + "grad_norm": 0.08407562226057053, + "learning_rate": 4.396614013712458e-06, + "loss": 0.0124, + "step": 91490 + }, + { + "epoch": 26.622054116962467, + "grad_norm": 0.08696401119232178, + "learning_rate": 4.393690581159399e-06, + "loss": 0.0118, + "step": 91500 + }, + { + "epoch": 26.624963631073612, + "grad_norm": 0.10573653131723404, + "learning_rate": 4.390767954085548e-06, + "loss": 0.0128, + "step": 91510 + }, + { + "epoch": 26.627873145184754, + "grad_norm": 0.09253575652837753, + "learning_rate": 4.3878461327128585e-06, + "loss": 0.0129, + "step": 91520 + }, + { + "epoch": 26.630782659295896, + "grad_norm": 0.09447692334651947, + "learning_rate": 4.38492511726322e-06, + "loss": 0.0114, + "step": 91530 + }, + { + "epoch": 26.63369217340704, + "grad_norm": 0.12340085208415985, + "learning_rate": 4.382004907958462e-06, + "loss": 0.0114, + "step": 91540 + }, + { + "epoch": 26.636601687518183, + "grad_norm": 0.1250937134027481, + "learning_rate": 4.37908550502036e-06, + "loss": 0.0101, + "step": 91550 + }, + { + "epoch": 26.63951120162933, + "grad_norm": 0.10686202347278595, + "learning_rate": 4.376166908670624e-06, + "loss": 0.0124, + "step": 91560 + }, + { + "epoch": 26.64242071574047, + "grad_norm": 0.13538463413715363, + "learning_rate": 4.373249119130897e-06, + "loss": 0.0106, + "step": 91570 + }, + { + "epoch": 26.645330229851616, + "grad_norm": 0.1337747424840927, + "learning_rate": 4.370332136622768e-06, + "loss": 0.0111, + "step": 91580 + }, + { + "epoch": 26.64823974396276, + "grad_norm": 0.09373465925455093, + "learning_rate": 4.367415961367769e-06, + "loss": 0.0113, + "step": 91590 + }, + { + "epoch": 26.6511492580739, + "grad_norm": 0.09710672497749329, + "learning_rate": 4.364500593587355e-06, + "loss": 0.0114, + "step": 91600 + }, + { + "epoch": 26.654058772185046, + "grad_norm": 0.12030233442783356, + "learning_rate": 4.361586033502936e-06, + "loss": 0.0103, + "step": 91610 + }, + { + "epoch": 26.656968286296188, + "grad_norm": 0.08214841037988663, + "learning_rate": 4.358672281335845e-06, + "loss": 0.0108, + "step": 91620 + }, + { + "epoch": 26.659877800407333, + "grad_norm": 0.08227057009935379, + "learning_rate": 4.355759337307374e-06, + "loss": 0.0104, + "step": 91630 + }, + { + "epoch": 26.662787314518475, + "grad_norm": 0.0918714627623558, + "learning_rate": 4.35284720163873e-06, + "loss": 0.0114, + "step": 91640 + }, + { + "epoch": 26.66569682862962, + "grad_norm": 0.16049595177173615, + "learning_rate": 4.34993587455108e-06, + "loss": 0.0133, + "step": 91650 + }, + { + "epoch": 26.668606342740762, + "grad_norm": 0.09874666482210159, + "learning_rate": 4.347025356265514e-06, + "loss": 0.0122, + "step": 91660 + }, + { + "epoch": 26.671515856851904, + "grad_norm": 0.08409792184829712, + "learning_rate": 4.3441156470030656e-06, + "loss": 0.0121, + "step": 91670 + }, + { + "epoch": 26.67442537096305, + "grad_norm": 0.09902283549308777, + "learning_rate": 4.341206746984717e-06, + "loss": 0.0112, + "step": 91680 + }, + { + "epoch": 26.677334885074192, + "grad_norm": 0.10682518780231476, + "learning_rate": 4.338298656431369e-06, + "loss": 0.0111, + "step": 91690 + }, + { + "epoch": 26.680244399185337, + "grad_norm": 0.09408320486545563, + "learning_rate": 4.335391375563876e-06, + "loss": 0.0108, + "step": 91700 + }, + { + "epoch": 26.68315391329648, + "grad_norm": 0.19932375848293304, + "learning_rate": 4.332484904603033e-06, + "loss": 0.0115, + "step": 91710 + }, + { + "epoch": 26.686063427407625, + "grad_norm": 0.2168942242860794, + "learning_rate": 4.329579243769564e-06, + "loss": 0.0116, + "step": 91720 + }, + { + "epoch": 26.688972941518767, + "grad_norm": 0.12121767550706863, + "learning_rate": 4.326674393284127e-06, + "loss": 0.0123, + "step": 91730 + }, + { + "epoch": 26.69188245562991, + "grad_norm": 0.13162629306316376, + "learning_rate": 4.3237703533673314e-06, + "loss": 0.0116, + "step": 91740 + }, + { + "epoch": 26.694791969741054, + "grad_norm": 0.13065341114997864, + "learning_rate": 4.320867124239726e-06, + "loss": 0.0111, + "step": 91750 + }, + { + "epoch": 26.697701483852196, + "grad_norm": 0.15767158567905426, + "learning_rate": 4.317964706121786e-06, + "loss": 0.0104, + "step": 91760 + }, + { + "epoch": 26.70061099796334, + "grad_norm": 0.12849470973014832, + "learning_rate": 4.3150630992339205e-06, + "loss": 0.0108, + "step": 91770 + }, + { + "epoch": 26.703520512074483, + "grad_norm": 0.11732284724712372, + "learning_rate": 4.312162303796509e-06, + "loss": 0.0119, + "step": 91780 + }, + { + "epoch": 26.706430026185625, + "grad_norm": 0.13909600675106049, + "learning_rate": 4.309262320029837e-06, + "loss": 0.0135, + "step": 91790 + }, + { + "epoch": 26.70933954029677, + "grad_norm": 0.1262485831975937, + "learning_rate": 4.3063631481541334e-06, + "loss": 0.0107, + "step": 91800 + }, + { + "epoch": 26.712249054407913, + "grad_norm": 0.09967502951622009, + "learning_rate": 4.3034647883895785e-06, + "loss": 0.0115, + "step": 91810 + }, + { + "epoch": 26.71515856851906, + "grad_norm": 0.0806422233581543, + "learning_rate": 4.3005672409562855e-06, + "loss": 0.0129, + "step": 91820 + }, + { + "epoch": 26.7180680826302, + "grad_norm": 0.11641861498355865, + "learning_rate": 4.297670506074299e-06, + "loss": 0.012, + "step": 91830 + }, + { + "epoch": 26.720977596741346, + "grad_norm": 0.10879762470722198, + "learning_rate": 4.2947745839636065e-06, + "loss": 0.0107, + "step": 91840 + }, + { + "epoch": 26.723887110852488, + "grad_norm": 0.12316453456878662, + "learning_rate": 4.291879474844135e-06, + "loss": 0.0101, + "step": 91850 + }, + { + "epoch": 26.72679662496363, + "grad_norm": 0.12228534370660782, + "learning_rate": 4.288985178935753e-06, + "loss": 0.0125, + "step": 91860 + }, + { + "epoch": 26.729706139074775, + "grad_norm": 0.06362485140562057, + "learning_rate": 4.286091696458257e-06, + "loss": 0.0098, + "step": 91870 + }, + { + "epoch": 26.732615653185917, + "grad_norm": 0.1321079283952713, + "learning_rate": 4.283199027631392e-06, + "loss": 0.0131, + "step": 91880 + }, + { + "epoch": 26.735525167297062, + "grad_norm": 0.10637672245502472, + "learning_rate": 4.280307172674839e-06, + "loss": 0.0114, + "step": 91890 + }, + { + "epoch": 26.738434681408204, + "grad_norm": 0.08570706844329834, + "learning_rate": 4.277416131808212e-06, + "loss": 0.0117, + "step": 91900 + }, + { + "epoch": 26.74134419551935, + "grad_norm": 0.10658394545316696, + "learning_rate": 4.2745259052510645e-06, + "loss": 0.0107, + "step": 91910 + }, + { + "epoch": 26.74425370963049, + "grad_norm": 0.06176143139600754, + "learning_rate": 4.27163649322289e-06, + "loss": 0.0114, + "step": 91920 + }, + { + "epoch": 26.747163223741634, + "grad_norm": 0.08122782409191132, + "learning_rate": 4.268747895943129e-06, + "loss": 0.0113, + "step": 91930 + }, + { + "epoch": 26.75007273785278, + "grad_norm": 0.10097743570804596, + "learning_rate": 4.265860113631142e-06, + "loss": 0.0105, + "step": 91940 + }, + { + "epoch": 26.75298225196392, + "grad_norm": 0.09584463387727737, + "learning_rate": 4.262973146506241e-06, + "loss": 0.0117, + "step": 91950 + }, + { + "epoch": 26.755891766075067, + "grad_norm": 0.11343914270401001, + "learning_rate": 4.260086994787669e-06, + "loss": 0.0111, + "step": 91960 + }, + { + "epoch": 26.75880128018621, + "grad_norm": 0.12820829451084137, + "learning_rate": 4.257201658694612e-06, + "loss": 0.0135, + "step": 91970 + }, + { + "epoch": 26.761710794297354, + "grad_norm": 0.1539735049009323, + "learning_rate": 4.2543171384461985e-06, + "loss": 0.0143, + "step": 91980 + }, + { + "epoch": 26.764620308408496, + "grad_norm": 0.11787393689155579, + "learning_rate": 4.251433434261477e-06, + "loss": 0.0106, + "step": 91990 + }, + { + "epoch": 26.767529822519638, + "grad_norm": 0.10539472848176956, + "learning_rate": 4.248550546359458e-06, + "loss": 0.011, + "step": 92000 + }, + { + "epoch": 26.770439336630783, + "grad_norm": 0.13080309331417084, + "learning_rate": 4.245668474959065e-06, + "loss": 0.0142, + "step": 92010 + }, + { + "epoch": 26.773348850741925, + "grad_norm": 0.26983627676963806, + "learning_rate": 4.242787220279185e-06, + "loss": 0.0125, + "step": 92020 + }, + { + "epoch": 26.77625836485307, + "grad_norm": 0.09890328347682953, + "learning_rate": 4.23990678253862e-06, + "loss": 0.0116, + "step": 92030 + }, + { + "epoch": 26.779167878964213, + "grad_norm": 0.11391404271125793, + "learning_rate": 4.237027161956122e-06, + "loss": 0.0117, + "step": 92040 + }, + { + "epoch": 26.782077393075355, + "grad_norm": 0.08783342689275742, + "learning_rate": 4.234148358750388e-06, + "loss": 0.0114, + "step": 92050 + }, + { + "epoch": 26.7849869071865, + "grad_norm": 0.1373830884695053, + "learning_rate": 4.231270373140037e-06, + "loss": 0.012, + "step": 92060 + }, + { + "epoch": 26.787896421297642, + "grad_norm": 0.11733800172805786, + "learning_rate": 4.228393205343626e-06, + "loss": 0.0135, + "step": 92070 + }, + { + "epoch": 26.790805935408788, + "grad_norm": 0.10977676510810852, + "learning_rate": 4.2255168555796715e-06, + "loss": 0.012, + "step": 92080 + }, + { + "epoch": 26.79371544951993, + "grad_norm": 0.08962979167699814, + "learning_rate": 4.2226413240666086e-06, + "loss": 0.011, + "step": 92090 + }, + { + "epoch": 26.796624963631075, + "grad_norm": 0.08938459306955338, + "learning_rate": 4.219766611022807e-06, + "loss": 0.0118, + "step": 92100 + }, + { + "epoch": 26.799534477742217, + "grad_norm": 0.0984835997223854, + "learning_rate": 4.216892716666588e-06, + "loss": 0.01, + "step": 92110 + }, + { + "epoch": 26.80244399185336, + "grad_norm": 0.10502730309963226, + "learning_rate": 4.214019641216209e-06, + "loss": 0.0118, + "step": 92120 + }, + { + "epoch": 26.805353505964504, + "grad_norm": 0.08048006147146225, + "learning_rate": 4.211147384889858e-06, + "loss": 0.0101, + "step": 92130 + }, + { + "epoch": 26.808263020075646, + "grad_norm": 0.11698340624570847, + "learning_rate": 4.208275947905659e-06, + "loss": 0.0123, + "step": 92140 + }, + { + "epoch": 26.81117253418679, + "grad_norm": 0.10749784857034683, + "learning_rate": 4.205405330481681e-06, + "loss": 0.0127, + "step": 92150 + }, + { + "epoch": 26.814082048297934, + "grad_norm": 0.12078143656253815, + "learning_rate": 4.202535532835934e-06, + "loss": 0.011, + "step": 92160 + }, + { + "epoch": 26.81699156240908, + "grad_norm": 0.11562138795852661, + "learning_rate": 4.199666555186354e-06, + "loss": 0.0127, + "step": 92170 + }, + { + "epoch": 26.81990107652022, + "grad_norm": 0.09818000346422195, + "learning_rate": 4.196798397750821e-06, + "loss": 0.0122, + "step": 92180 + }, + { + "epoch": 26.822810590631363, + "grad_norm": 0.12329849600791931, + "learning_rate": 4.1939310607471595e-06, + "loss": 0.0121, + "step": 92190 + }, + { + "epoch": 26.82572010474251, + "grad_norm": 0.08824770897626877, + "learning_rate": 4.191064544393122e-06, + "loss": 0.0144, + "step": 92200 + }, + { + "epoch": 26.82862961885365, + "grad_norm": 0.08534722030162811, + "learning_rate": 4.188198848906393e-06, + "loss": 0.0101, + "step": 92210 + }, + { + "epoch": 26.831539132964796, + "grad_norm": 0.14103595912456512, + "learning_rate": 4.185333974504611e-06, + "loss": 0.0112, + "step": 92220 + }, + { + "epoch": 26.834448647075938, + "grad_norm": 0.09276002645492554, + "learning_rate": 4.1824699214053474e-06, + "loss": 0.0102, + "step": 92230 + }, + { + "epoch": 26.837358161187083, + "grad_norm": 0.13936731219291687, + "learning_rate": 4.179606689826098e-06, + "loss": 0.0115, + "step": 92240 + }, + { + "epoch": 26.840267675298225, + "grad_norm": 0.12857003509998322, + "learning_rate": 4.176744279984319e-06, + "loss": 0.0123, + "step": 92250 + }, + { + "epoch": 26.843177189409367, + "grad_norm": 0.10333877801895142, + "learning_rate": 4.173882692097379e-06, + "loss": 0.0108, + "step": 92260 + }, + { + "epoch": 26.846086703520513, + "grad_norm": 0.08836212754249573, + "learning_rate": 4.171021926382605e-06, + "loss": 0.0095, + "step": 92270 + }, + { + "epoch": 26.848996217631655, + "grad_norm": 0.10284505039453506, + "learning_rate": 4.1681619830572484e-06, + "loss": 0.0116, + "step": 92280 + }, + { + "epoch": 26.8519057317428, + "grad_norm": 0.11580946296453476, + "learning_rate": 4.165302862338504e-06, + "loss": 0.0098, + "step": 92290 + }, + { + "epoch": 26.854815245853942, + "grad_norm": 0.1068500280380249, + "learning_rate": 4.1624445644435094e-06, + "loss": 0.0145, + "step": 92300 + }, + { + "epoch": 26.857724759965087, + "grad_norm": 0.473379909992218, + "learning_rate": 4.159587089589325e-06, + "loss": 0.0108, + "step": 92310 + }, + { + "epoch": 26.86063427407623, + "grad_norm": 0.11877524852752686, + "learning_rate": 4.156730437992965e-06, + "loss": 0.0114, + "step": 92320 + }, + { + "epoch": 26.86354378818737, + "grad_norm": 0.10467761754989624, + "learning_rate": 4.153874609871364e-06, + "loss": 0.0109, + "step": 92330 + }, + { + "epoch": 26.866453302298517, + "grad_norm": 0.13698382675647736, + "learning_rate": 4.151019605441406e-06, + "loss": 0.0119, + "step": 92340 + }, + { + "epoch": 26.86936281640966, + "grad_norm": 0.095960333943367, + "learning_rate": 4.148165424919919e-06, + "loss": 0.0122, + "step": 92350 + }, + { + "epoch": 26.872272330520804, + "grad_norm": 0.22270652651786804, + "learning_rate": 4.145312068523652e-06, + "loss": 0.011, + "step": 92360 + }, + { + "epoch": 26.875181844631946, + "grad_norm": 0.07740391790866852, + "learning_rate": 4.142459536469292e-06, + "loss": 0.0108, + "step": 92370 + }, + { + "epoch": 26.87809135874309, + "grad_norm": 0.07714121043682098, + "learning_rate": 4.139607828973479e-06, + "loss": 0.0123, + "step": 92380 + }, + { + "epoch": 26.881000872854234, + "grad_norm": 0.1104799434542656, + "learning_rate": 4.136756946252784e-06, + "loss": 0.0116, + "step": 92390 + }, + { + "epoch": 26.883910386965375, + "grad_norm": 0.1066170334815979, + "learning_rate": 4.133906888523702e-06, + "loss": 0.0121, + "step": 92400 + }, + { + "epoch": 26.88681990107652, + "grad_norm": 0.09247691184282303, + "learning_rate": 4.1310576560026816e-06, + "loss": 0.0117, + "step": 92410 + }, + { + "epoch": 26.889729415187663, + "grad_norm": 0.09868945181369781, + "learning_rate": 4.128209248906108e-06, + "loss": 0.0123, + "step": 92420 + }, + { + "epoch": 26.89263892929881, + "grad_norm": 0.10409216582775116, + "learning_rate": 4.125361667450295e-06, + "loss": 0.013, + "step": 92430 + }, + { + "epoch": 26.89554844340995, + "grad_norm": 0.09770956635475159, + "learning_rate": 4.122514911851491e-06, + "loss": 0.0138, + "step": 92440 + }, + { + "epoch": 26.898457957521096, + "grad_norm": 0.10641931742429733, + "learning_rate": 4.119668982325897e-06, + "loss": 0.0118, + "step": 92450 + }, + { + "epoch": 26.901367471632238, + "grad_norm": 0.11176233738660812, + "learning_rate": 4.116823879089642e-06, + "loss": 0.0097, + "step": 92460 + }, + { + "epoch": 26.90427698574338, + "grad_norm": 0.1121622622013092, + "learning_rate": 4.113979602358792e-06, + "loss": 0.0123, + "step": 92470 + }, + { + "epoch": 26.907186499854525, + "grad_norm": 0.10391731560230255, + "learning_rate": 4.11113615234934e-06, + "loss": 0.011, + "step": 92480 + }, + { + "epoch": 26.910096013965667, + "grad_norm": 0.08893290162086487, + "learning_rate": 4.108293529277245e-06, + "loss": 0.0109, + "step": 92490 + }, + { + "epoch": 26.913005528076813, + "grad_norm": 0.1407746821641922, + "learning_rate": 4.105451733358379e-06, + "loss": 0.0113, + "step": 92500 + }, + { + "epoch": 26.915915042187954, + "grad_norm": 0.09385742247104645, + "learning_rate": 4.102610764808551e-06, + "loss": 0.0108, + "step": 92510 + }, + { + "epoch": 26.918824556299096, + "grad_norm": 0.1335674226284027, + "learning_rate": 4.099770623843519e-06, + "loss": 0.0107, + "step": 92520 + }, + { + "epoch": 26.921734070410242, + "grad_norm": 0.07395434379577637, + "learning_rate": 4.096931310678977e-06, + "loss": 0.0112, + "step": 92530 + }, + { + "epoch": 26.924643584521384, + "grad_norm": 0.11884579062461853, + "learning_rate": 4.094092825530544e-06, + "loss": 0.0135, + "step": 92540 + }, + { + "epoch": 26.92755309863253, + "grad_norm": 0.08266336470842361, + "learning_rate": 4.091255168613792e-06, + "loss": 0.0122, + "step": 92550 + }, + { + "epoch": 26.93046261274367, + "grad_norm": 0.10173249989748001, + "learning_rate": 4.0884183401442155e-06, + "loss": 0.0117, + "step": 92560 + }, + { + "epoch": 26.933372126854817, + "grad_norm": 0.1460864245891571, + "learning_rate": 4.08558234033726e-06, + "loss": 0.012, + "step": 92570 + }, + { + "epoch": 26.93628164096596, + "grad_norm": 0.07080257683992386, + "learning_rate": 4.082747169408292e-06, + "loss": 0.0101, + "step": 92580 + }, + { + "epoch": 26.9391911550771, + "grad_norm": 0.10333485156297684, + "learning_rate": 4.079912827572629e-06, + "loss": 0.0131, + "step": 92590 + }, + { + "epoch": 26.942100669188246, + "grad_norm": 0.0816969946026802, + "learning_rate": 4.077079315045524e-06, + "loss": 0.0091, + "step": 92600 + }, + { + "epoch": 26.945010183299388, + "grad_norm": 0.15164759755134583, + "learning_rate": 4.0742466320421565e-06, + "loss": 0.012, + "step": 92610 + }, + { + "epoch": 26.947919697410534, + "grad_norm": 0.1259452998638153, + "learning_rate": 4.071414778777657e-06, + "loss": 0.0098, + "step": 92620 + }, + { + "epoch": 26.950829211521675, + "grad_norm": 0.11392171680927277, + "learning_rate": 4.06858375546708e-06, + "loss": 0.0108, + "step": 92630 + }, + { + "epoch": 26.95373872563282, + "grad_norm": 0.06921622902154922, + "learning_rate": 4.065753562325428e-06, + "loss": 0.0107, + "step": 92640 + }, + { + "epoch": 26.956648239743963, + "grad_norm": 0.19174450635910034, + "learning_rate": 4.062924199567631e-06, + "loss": 0.0121, + "step": 92650 + }, + { + "epoch": 26.959557753855105, + "grad_norm": 0.12076525390148163, + "learning_rate": 4.060095667408566e-06, + "loss": 0.0114, + "step": 92660 + }, + { + "epoch": 26.96246726796625, + "grad_norm": 0.10767622292041779, + "learning_rate": 4.057267966063034e-06, + "loss": 0.0125, + "step": 92670 + }, + { + "epoch": 26.965376782077392, + "grad_norm": 0.15069574117660522, + "learning_rate": 4.054441095745782e-06, + "loss": 0.012, + "step": 92680 + }, + { + "epoch": 26.968286296188538, + "grad_norm": 0.14754079282283783, + "learning_rate": 4.0516150566715005e-06, + "loss": 0.0115, + "step": 92690 + }, + { + "epoch": 26.97119581029968, + "grad_norm": 0.09692368656396866, + "learning_rate": 4.048789849054798e-06, + "loss": 0.0118, + "step": 92700 + }, + { + "epoch": 26.974105324410825, + "grad_norm": 0.10250865668058395, + "learning_rate": 4.045965473110235e-06, + "loss": 0.0125, + "step": 92710 + }, + { + "epoch": 26.977014838521967, + "grad_norm": 0.1325760781764984, + "learning_rate": 4.043141929052309e-06, + "loss": 0.0132, + "step": 92720 + }, + { + "epoch": 26.97992435263311, + "grad_norm": 0.0843382179737091, + "learning_rate": 4.0403192170954435e-06, + "loss": 0.0134, + "step": 92730 + }, + { + "epoch": 26.982833866744254, + "grad_norm": 0.11270768195390701, + "learning_rate": 4.037497337454001e-06, + "loss": 0.0115, + "step": 92740 + }, + { + "epoch": 26.985743380855396, + "grad_norm": 0.18487466871738434, + "learning_rate": 4.034676290342289e-06, + "loss": 0.0108, + "step": 92750 + }, + { + "epoch": 26.988652894966542, + "grad_norm": 0.10621064901351929, + "learning_rate": 4.031856075974555e-06, + "loss": 0.0118, + "step": 92760 + }, + { + "epoch": 26.991562409077684, + "grad_norm": 0.10340508818626404, + "learning_rate": 4.029036694564968e-06, + "loss": 0.0115, + "step": 92770 + }, + { + "epoch": 26.994471923188826, + "grad_norm": 0.08859309554100037, + "learning_rate": 4.0262181463276316e-06, + "loss": 0.0097, + "step": 92780 + }, + { + "epoch": 26.99738143729997, + "grad_norm": 0.1010841578245163, + "learning_rate": 4.023400431476615e-06, + "loss": 0.0123, + "step": 92790 + }, + { + "epoch": 27.000290951411113, + "grad_norm": 0.0801166519522667, + "learning_rate": 4.020583550225896e-06, + "loss": 0.013, + "step": 92800 + }, + { + "epoch": 27.00320046552226, + "grad_norm": 0.17988455295562744, + "learning_rate": 4.0177675027893965e-06, + "loss": 0.0114, + "step": 92810 + }, + { + "epoch": 27.0061099796334, + "grad_norm": 0.12971754372119904, + "learning_rate": 4.014952289380977e-06, + "loss": 0.0122, + "step": 92820 + }, + { + "epoch": 27.009019493744546, + "grad_norm": 0.09873839467763901, + "learning_rate": 4.012137910214441e-06, + "loss": 0.0106, + "step": 92830 + }, + { + "epoch": 27.011929007855688, + "grad_norm": 0.08898330479860306, + "learning_rate": 4.009324365503516e-06, + "loss": 0.0107, + "step": 92840 + }, + { + "epoch": 27.01483852196683, + "grad_norm": 0.1646529585123062, + "learning_rate": 4.006511655461871e-06, + "loss": 0.0122, + "step": 92850 + }, + { + "epoch": 27.017748036077975, + "grad_norm": 0.09366992861032486, + "learning_rate": 4.003699780303114e-06, + "loss": 0.0133, + "step": 92860 + }, + { + "epoch": 27.020657550189117, + "grad_norm": 0.08683478832244873, + "learning_rate": 4.000888740240793e-06, + "loss": 0.011, + "step": 92870 + }, + { + "epoch": 27.023567064300263, + "grad_norm": 0.29735320806503296, + "learning_rate": 3.998078535488382e-06, + "loss": 0.0117, + "step": 92880 + }, + { + "epoch": 27.026476578411405, + "grad_norm": 0.1347411721944809, + "learning_rate": 3.9952691662592995e-06, + "loss": 0.0114, + "step": 92890 + }, + { + "epoch": 27.02938609252255, + "grad_norm": 0.08631578832864761, + "learning_rate": 3.992460632766902e-06, + "loss": 0.0111, + "step": 92900 + }, + { + "epoch": 27.032295606633692, + "grad_norm": 0.1330031305551529, + "learning_rate": 3.9896529352244715e-06, + "loss": 0.0137, + "step": 92910 + }, + { + "epoch": 27.035205120744834, + "grad_norm": 0.08556458353996277, + "learning_rate": 3.986846073845243e-06, + "loss": 0.0129, + "step": 92920 + }, + { + "epoch": 27.03811463485598, + "grad_norm": 0.17518799006938934, + "learning_rate": 3.984040048842372e-06, + "loss": 0.0127, + "step": 92930 + }, + { + "epoch": 27.04102414896712, + "grad_norm": 0.12118197232484818, + "learning_rate": 3.981234860428963e-06, + "loss": 0.0104, + "step": 92940 + }, + { + "epoch": 27.043933663078267, + "grad_norm": 0.11654990911483765, + "learning_rate": 3.978430508818044e-06, + "loss": 0.0122, + "step": 92950 + }, + { + "epoch": 27.04684317718941, + "grad_norm": 0.09441301971673965, + "learning_rate": 3.9756269942225985e-06, + "loss": 0.0124, + "step": 92960 + }, + { + "epoch": 27.049752691300554, + "grad_norm": 0.12860102951526642, + "learning_rate": 3.972824316855524e-06, + "loss": 0.0099, + "step": 92970 + }, + { + "epoch": 27.052662205411696, + "grad_norm": 0.09998370707035065, + "learning_rate": 3.970022476929667e-06, + "loss": 0.0113, + "step": 92980 + }, + { + "epoch": 27.05557171952284, + "grad_norm": 0.13079224526882172, + "learning_rate": 3.9672214746578165e-06, + "loss": 0.0133, + "step": 92990 + }, + { + "epoch": 27.058481233633984, + "grad_norm": 0.07109879702329636, + "learning_rate": 3.964421310252682e-06, + "loss": 0.0125, + "step": 93000 + }, + { + "epoch": 27.061390747745126, + "grad_norm": 0.09731003642082214, + "learning_rate": 3.961621983926923e-06, + "loss": 0.0116, + "step": 93010 + }, + { + "epoch": 27.06430026185627, + "grad_norm": 0.09186958521604538, + "learning_rate": 3.958823495893124e-06, + "loss": 0.0116, + "step": 93020 + }, + { + "epoch": 27.067209775967413, + "grad_norm": 0.15172959864139557, + "learning_rate": 3.95602584636382e-06, + "loss": 0.0108, + "step": 93030 + }, + { + "epoch": 27.07011929007856, + "grad_norm": 0.09289747476577759, + "learning_rate": 3.953229035551464e-06, + "loss": 0.0117, + "step": 93040 + }, + { + "epoch": 27.0730288041897, + "grad_norm": 0.07643039524555206, + "learning_rate": 3.95043306366846e-06, + "loss": 0.0109, + "step": 93050 + }, + { + "epoch": 27.075938318300842, + "grad_norm": 0.10168536752462387, + "learning_rate": 3.94763793092715e-06, + "loss": 0.0125, + "step": 93060 + }, + { + "epoch": 27.078847832411988, + "grad_norm": 0.09128133952617645, + "learning_rate": 3.944843637539797e-06, + "loss": 0.0116, + "step": 93070 + }, + { + "epoch": 27.08175734652313, + "grad_norm": 0.08423581719398499, + "learning_rate": 3.942050183718606e-06, + "loss": 0.0113, + "step": 93080 + }, + { + "epoch": 27.084666860634275, + "grad_norm": 0.09448714554309845, + "learning_rate": 3.939257569675735e-06, + "loss": 0.0107, + "step": 93090 + }, + { + "epoch": 27.087576374745417, + "grad_norm": 0.13805058598518372, + "learning_rate": 3.936465795623257e-06, + "loss": 0.0112, + "step": 93100 + }, + { + "epoch": 27.090485888856563, + "grad_norm": 0.11928583681583405, + "learning_rate": 3.933674861773185e-06, + "loss": 0.0118, + "step": 93110 + }, + { + "epoch": 27.093395402967705, + "grad_norm": 0.12881624698638916, + "learning_rate": 3.930884768337476e-06, + "loss": 0.0134, + "step": 93120 + }, + { + "epoch": 27.096304917078847, + "grad_norm": 0.09364187717437744, + "learning_rate": 3.928095515528022e-06, + "loss": 0.0107, + "step": 93130 + }, + { + "epoch": 27.099214431189992, + "grad_norm": 0.34514328837394714, + "learning_rate": 3.9253071035566476e-06, + "loss": 0.0106, + "step": 93140 + }, + { + "epoch": 27.102123945301134, + "grad_norm": 0.08505402505397797, + "learning_rate": 3.9225195326351056e-06, + "loss": 0.0109, + "step": 93150 + }, + { + "epoch": 27.10503345941228, + "grad_norm": 0.13663069903850555, + "learning_rate": 3.919732802975102e-06, + "loss": 0.012, + "step": 93160 + }, + { + "epoch": 27.10794297352342, + "grad_norm": 0.12823836505413055, + "learning_rate": 3.916946914788272e-06, + "loss": 0.012, + "step": 93170 + }, + { + "epoch": 27.110852487634563, + "grad_norm": 0.127447247505188, + "learning_rate": 3.914161868286177e-06, + "loss": 0.01, + "step": 93180 + }, + { + "epoch": 27.11376200174571, + "grad_norm": 0.12690603733062744, + "learning_rate": 3.9113776636803296e-06, + "loss": 0.0112, + "step": 93190 + }, + { + "epoch": 27.11667151585685, + "grad_norm": 0.09973982721567154, + "learning_rate": 3.908594301182172e-06, + "loss": 0.0109, + "step": 93200 + }, + { + "epoch": 27.119581029967996, + "grad_norm": 0.1418268382549286, + "learning_rate": 3.905811781003083e-06, + "loss": 0.0107, + "step": 93210 + }, + { + "epoch": 27.122490544079138, + "grad_norm": 0.12688018381595612, + "learning_rate": 3.9030301033543674e-06, + "loss": 0.0102, + "step": 93220 + }, + { + "epoch": 27.125400058190284, + "grad_norm": 0.1264631599187851, + "learning_rate": 3.900249268447283e-06, + "loss": 0.0106, + "step": 93230 + }, + { + "epoch": 27.128309572301426, + "grad_norm": 0.10690052062273026, + "learning_rate": 3.897469276493018e-06, + "loss": 0.0109, + "step": 93240 + }, + { + "epoch": 27.131219086412568, + "grad_norm": 0.12037681043148041, + "learning_rate": 3.894690127702689e-06, + "loss": 0.0113, + "step": 93250 + }, + { + "epoch": 27.134128600523713, + "grad_norm": 0.11801845580339432, + "learning_rate": 3.891911822287359e-06, + "loss": 0.0102, + "step": 93260 + }, + { + "epoch": 27.137038114634855, + "grad_norm": 0.10939665138721466, + "learning_rate": 3.889134360458014e-06, + "loss": 0.0096, + "step": 93270 + }, + { + "epoch": 27.139947628746, + "grad_norm": 0.08864489197731018, + "learning_rate": 3.88635774242559e-06, + "loss": 0.0139, + "step": 93280 + }, + { + "epoch": 27.142857142857142, + "grad_norm": 0.07263367623090744, + "learning_rate": 3.883581968400957e-06, + "loss": 0.0111, + "step": 93290 + }, + { + "epoch": 27.145766656968288, + "grad_norm": 0.15792590379714966, + "learning_rate": 3.8808070385949065e-06, + "loss": 0.0102, + "step": 93300 + }, + { + "epoch": 27.14867617107943, + "grad_norm": 0.09783598780632019, + "learning_rate": 3.878032953218183e-06, + "loss": 0.0112, + "step": 93310 + }, + { + "epoch": 27.15158568519057, + "grad_norm": 0.0786840096116066, + "learning_rate": 3.875259712481457e-06, + "loss": 0.0143, + "step": 93320 + }, + { + "epoch": 27.154495199301717, + "grad_norm": 0.13996651768684387, + "learning_rate": 3.872487316595342e-06, + "loss": 0.0123, + "step": 93330 + }, + { + "epoch": 27.15740471341286, + "grad_norm": 0.15401732921600342, + "learning_rate": 3.8697157657703755e-06, + "loss": 0.0137, + "step": 93340 + }, + { + "epoch": 27.160314227524005, + "grad_norm": 0.10242515057325363, + "learning_rate": 3.866945060217044e-06, + "loss": 0.0117, + "step": 93350 + }, + { + "epoch": 27.163223741635147, + "grad_norm": 0.11108355969190598, + "learning_rate": 3.864175200145767e-06, + "loss": 0.0144, + "step": 93360 + }, + { + "epoch": 27.166133255746292, + "grad_norm": 0.08687623590230942, + "learning_rate": 3.861406185766894e-06, + "loss": 0.0095, + "step": 93370 + }, + { + "epoch": 27.169042769857434, + "grad_norm": 0.07990217208862305, + "learning_rate": 3.858638017290704e-06, + "loss": 0.0102, + "step": 93380 + }, + { + "epoch": 27.171952283968576, + "grad_norm": 0.10726264119148254, + "learning_rate": 3.855870694927439e-06, + "loss": 0.0094, + "step": 93390 + }, + { + "epoch": 27.17486179807972, + "grad_norm": 0.09072327613830566, + "learning_rate": 3.853104218887249e-06, + "loss": 0.0114, + "step": 93400 + }, + { + "epoch": 27.177771312190863, + "grad_norm": 0.09515589475631714, + "learning_rate": 3.8503385893802275e-06, + "loss": 0.0111, + "step": 93410 + }, + { + "epoch": 27.18068082630201, + "grad_norm": 0.10817790031433105, + "learning_rate": 3.847573806616408e-06, + "loss": 0.0115, + "step": 93420 + }, + { + "epoch": 27.18359034041315, + "grad_norm": 0.13244600594043732, + "learning_rate": 3.8448098708057626e-06, + "loss": 0.0156, + "step": 93430 + }, + { + "epoch": 27.186499854524296, + "grad_norm": 0.12518276274204254, + "learning_rate": 3.842046782158189e-06, + "loss": 0.0108, + "step": 93440 + }, + { + "epoch": 27.189409368635438, + "grad_norm": 0.12219314277172089, + "learning_rate": 3.839284540883522e-06, + "loss": 0.0127, + "step": 93450 + }, + { + "epoch": 27.19231888274658, + "grad_norm": 0.15441487729549408, + "learning_rate": 3.836523147191541e-06, + "loss": 0.0113, + "step": 93460 + }, + { + "epoch": 27.195228396857726, + "grad_norm": 0.06515106558799744, + "learning_rate": 3.833762601291958e-06, + "loss": 0.0116, + "step": 93470 + }, + { + "epoch": 27.198137910968867, + "grad_norm": 0.0819002017378807, + "learning_rate": 3.831002903394414e-06, + "loss": 0.0119, + "step": 93480 + }, + { + "epoch": 27.201047425080013, + "grad_norm": 0.11505831032991409, + "learning_rate": 3.828244053708484e-06, + "loss": 0.012, + "step": 93490 + }, + { + "epoch": 27.203956939191155, + "grad_norm": 0.11051633954048157, + "learning_rate": 3.825486052443697e-06, + "loss": 0.0095, + "step": 93500 + }, + { + "epoch": 27.206866453302297, + "grad_norm": 0.09388643503189087, + "learning_rate": 3.822728899809501e-06, + "loss": 0.0132, + "step": 93510 + }, + { + "epoch": 27.209775967413442, + "grad_norm": 0.10651973634958267, + "learning_rate": 3.819972596015276e-06, + "loss": 0.0093, + "step": 93520 + }, + { + "epoch": 27.212685481524584, + "grad_norm": 0.08555954694747925, + "learning_rate": 3.817217141270351e-06, + "loss": 0.01, + "step": 93530 + }, + { + "epoch": 27.21559499563573, + "grad_norm": 0.11285668611526489, + "learning_rate": 3.8144625357839885e-06, + "loss": 0.0112, + "step": 93540 + }, + { + "epoch": 27.21850450974687, + "grad_norm": 0.09565701335668564, + "learning_rate": 3.8117087797653727e-06, + "loss": 0.0123, + "step": 93550 + }, + { + "epoch": 27.221414023858017, + "grad_norm": 0.10066480934619904, + "learning_rate": 3.808955873423646e-06, + "loss": 0.0128, + "step": 93560 + }, + { + "epoch": 27.22432353796916, + "grad_norm": 0.17633379995822906, + "learning_rate": 3.8062038169678588e-06, + "loss": 0.0127, + "step": 93570 + }, + { + "epoch": 27.2272330520803, + "grad_norm": 0.09696795791387558, + "learning_rate": 3.8034526106070244e-06, + "loss": 0.0111, + "step": 93580 + }, + { + "epoch": 27.230142566191446, + "grad_norm": 0.13921082019805908, + "learning_rate": 3.8007022545500706e-06, + "loss": 0.0109, + "step": 93590 + }, + { + "epoch": 27.23305208030259, + "grad_norm": 0.10837170481681824, + "learning_rate": 3.7979527490058707e-06, + "loss": 0.0111, + "step": 93600 + }, + { + "epoch": 27.235961594413734, + "grad_norm": 0.14415352046489716, + "learning_rate": 3.7952040941832356e-06, + "loss": 0.0087, + "step": 93610 + }, + { + "epoch": 27.238871108524876, + "grad_norm": 0.13389325141906738, + "learning_rate": 3.7924562902909018e-06, + "loss": 0.0125, + "step": 93620 + }, + { + "epoch": 27.24178062263602, + "grad_norm": 0.09202083200216293, + "learning_rate": 3.789709337537552e-06, + "loss": 0.0116, + "step": 93630 + }, + { + "epoch": 27.244690136747163, + "grad_norm": 0.07119566202163696, + "learning_rate": 3.7869632361317928e-06, + "loss": 0.012, + "step": 93640 + }, + { + "epoch": 27.247599650858305, + "grad_norm": 0.07909061014652252, + "learning_rate": 3.784217986282178e-06, + "loss": 0.0104, + "step": 93650 + }, + { + "epoch": 27.25050916496945, + "grad_norm": 0.11125337332487106, + "learning_rate": 3.781473588197192e-06, + "loss": 0.01, + "step": 93660 + }, + { + "epoch": 27.253418679080593, + "grad_norm": 0.11387079209089279, + "learning_rate": 3.778730042085251e-06, + "loss": 0.0115, + "step": 93670 + }, + { + "epoch": 27.256328193191738, + "grad_norm": 0.08182413130998611, + "learning_rate": 3.7759873481547057e-06, + "loss": 0.0123, + "step": 93680 + }, + { + "epoch": 27.25923770730288, + "grad_norm": 0.08462554961442947, + "learning_rate": 3.7732455066138493e-06, + "loss": 0.0132, + "step": 93690 + }, + { + "epoch": 27.262147221414025, + "grad_norm": 0.1540762186050415, + "learning_rate": 3.77050451767091e-06, + "loss": 0.0121, + "step": 93700 + }, + { + "epoch": 27.265056735525167, + "grad_norm": 0.1305098682641983, + "learning_rate": 3.7677643815340403e-06, + "loss": 0.0106, + "step": 93710 + }, + { + "epoch": 27.26796624963631, + "grad_norm": 0.08059406280517578, + "learning_rate": 3.7650250984113393e-06, + "loss": 0.0114, + "step": 93720 + }, + { + "epoch": 27.270875763747455, + "grad_norm": 0.12834946811199188, + "learning_rate": 3.7622866685108433e-06, + "loss": 0.0111, + "step": 93730 + }, + { + "epoch": 27.273785277858597, + "grad_norm": 0.14321938157081604, + "learning_rate": 3.759549092040512e-06, + "loss": 0.012, + "step": 93740 + }, + { + "epoch": 27.276694791969742, + "grad_norm": 0.11801843345165253, + "learning_rate": 3.7568123692082423e-06, + "loss": 0.0114, + "step": 93750 + }, + { + "epoch": 27.279604306080884, + "grad_norm": 0.1255379319190979, + "learning_rate": 3.7540765002218736e-06, + "loss": 0.0124, + "step": 93760 + }, + { + "epoch": 27.28251382019203, + "grad_norm": 0.1803712099790573, + "learning_rate": 3.7513414852891826e-06, + "loss": 0.011, + "step": 93770 + }, + { + "epoch": 27.28542333430317, + "grad_norm": 0.0934293270111084, + "learning_rate": 3.7486073246178714e-06, + "loss": 0.0109, + "step": 93780 + }, + { + "epoch": 27.288332848414314, + "grad_norm": 0.09717763215303421, + "learning_rate": 3.7458740184155726e-06, + "loss": 0.011, + "step": 93790 + }, + { + "epoch": 27.29124236252546, + "grad_norm": 0.10298573225736618, + "learning_rate": 3.74314156688988e-06, + "loss": 0.011, + "step": 93800 + }, + { + "epoch": 27.2941518766366, + "grad_norm": 0.06576182693243027, + "learning_rate": 3.7404099702482973e-06, + "loss": 0.009, + "step": 93810 + }, + { + "epoch": 27.297061390747746, + "grad_norm": 0.10978743433952332, + "learning_rate": 3.737679228698266e-06, + "loss": 0.0112, + "step": 93820 + }, + { + "epoch": 27.29997090485889, + "grad_norm": 0.0954269990324974, + "learning_rate": 3.7349493424471708e-06, + "loss": 0.013, + "step": 93830 + }, + { + "epoch": 27.30288041897003, + "grad_norm": 0.13360445201396942, + "learning_rate": 3.732220311702337e-06, + "loss": 0.0113, + "step": 93840 + }, + { + "epoch": 27.305789933081176, + "grad_norm": 0.07584881037473679, + "learning_rate": 3.729492136671008e-06, + "loss": 0.0116, + "step": 93850 + }, + { + "epoch": 27.308699447192318, + "grad_norm": 0.1658412218093872, + "learning_rate": 3.726764817560368e-06, + "loss": 0.0108, + "step": 93860 + }, + { + "epoch": 27.311608961303463, + "grad_norm": 0.0883476734161377, + "learning_rate": 3.7240383545775437e-06, + "loss": 0.0098, + "step": 93870 + }, + { + "epoch": 27.314518475414605, + "grad_norm": 0.09576579183340073, + "learning_rate": 3.7213127479295954e-06, + "loss": 0.0112, + "step": 93880 + }, + { + "epoch": 27.31742798952575, + "grad_norm": 0.09901919215917587, + "learning_rate": 3.7185879978235064e-06, + "loss": 0.0118, + "step": 93890 + }, + { + "epoch": 27.320337503636893, + "grad_norm": 0.07786305993795395, + "learning_rate": 3.7158641044662096e-06, + "loss": 0.0114, + "step": 93900 + }, + { + "epoch": 27.323247017748034, + "grad_norm": 0.09601613879203796, + "learning_rate": 3.713141068064569e-06, + "loss": 0.013, + "step": 93910 + }, + { + "epoch": 27.32615653185918, + "grad_norm": 0.10581165552139282, + "learning_rate": 3.7104188888253726e-06, + "loss": 0.0107, + "step": 93920 + }, + { + "epoch": 27.329066045970322, + "grad_norm": 0.14208471775054932, + "learning_rate": 3.7076975669553604e-06, + "loss": 0.0121, + "step": 93930 + }, + { + "epoch": 27.331975560081467, + "grad_norm": 0.1389925628900528, + "learning_rate": 3.7049771026611916e-06, + "loss": 0.0127, + "step": 93940 + }, + { + "epoch": 27.33488507419261, + "grad_norm": 0.1001415029168129, + "learning_rate": 3.7022574961494764e-06, + "loss": 0.0106, + "step": 93950 + }, + { + "epoch": 27.337794588303755, + "grad_norm": 0.12371004372835159, + "learning_rate": 3.6995387476267412e-06, + "loss": 0.0101, + "step": 93960 + }, + { + "epoch": 27.340704102414897, + "grad_norm": 0.1526404768228531, + "learning_rate": 3.6968208572994656e-06, + "loss": 0.0112, + "step": 93970 + }, + { + "epoch": 27.34361361652604, + "grad_norm": 0.05377037078142166, + "learning_rate": 3.694103825374046e-06, + "loss": 0.0109, + "step": 93980 + }, + { + "epoch": 27.346523130637184, + "grad_norm": 0.08587592095136642, + "learning_rate": 3.6913876520568296e-06, + "loss": 0.0108, + "step": 93990 + }, + { + "epoch": 27.349432644748326, + "grad_norm": 0.11606904864311218, + "learning_rate": 3.6886723375540958e-06, + "loss": 0.0115, + "step": 94000 + }, + { + "epoch": 27.35234215885947, + "grad_norm": 0.09281585365533829, + "learning_rate": 3.6859578820720434e-06, + "loss": 0.0105, + "step": 94010 + }, + { + "epoch": 27.355251672970613, + "grad_norm": 0.10341320186853409, + "learning_rate": 3.683244285816826e-06, + "loss": 0.011, + "step": 94020 + }, + { + "epoch": 27.35816118708176, + "grad_norm": 0.08330871164798737, + "learning_rate": 3.6805315489945245e-06, + "loss": 0.0096, + "step": 94030 + }, + { + "epoch": 27.3610707011929, + "grad_norm": 0.09992986917495728, + "learning_rate": 3.6778196718111485e-06, + "loss": 0.0106, + "step": 94040 + }, + { + "epoch": 27.363980215304043, + "grad_norm": 0.10565102845430374, + "learning_rate": 3.6751086544726457e-06, + "loss": 0.0155, + "step": 94050 + }, + { + "epoch": 27.36688972941519, + "grad_norm": 0.14161178469657898, + "learning_rate": 3.6723984971849024e-06, + "loss": 0.0126, + "step": 94060 + }, + { + "epoch": 27.36979924352633, + "grad_norm": 0.09731929749250412, + "learning_rate": 3.669689200153741e-06, + "loss": 0.0126, + "step": 94070 + }, + { + "epoch": 27.372708757637476, + "grad_norm": 0.09061943739652634, + "learning_rate": 3.666980763584912e-06, + "loss": 0.0123, + "step": 94080 + }, + { + "epoch": 27.375618271748618, + "grad_norm": 0.10220864415168762, + "learning_rate": 3.664273187684094e-06, + "loss": 0.0107, + "step": 94090 + }, + { + "epoch": 27.378527785859763, + "grad_norm": 0.07552880793809891, + "learning_rate": 3.661566472656927e-06, + "loss": 0.0134, + "step": 94100 + }, + { + "epoch": 27.381437299970905, + "grad_norm": 0.09306006133556366, + "learning_rate": 3.6588606187089597e-06, + "loss": 0.0108, + "step": 94110 + }, + { + "epoch": 27.384346814082047, + "grad_norm": 0.10749334841966629, + "learning_rate": 3.6561556260456775e-06, + "loss": 0.0098, + "step": 94120 + }, + { + "epoch": 27.387256328193192, + "grad_norm": 0.11530201882123947, + "learning_rate": 3.6534514948725145e-06, + "loss": 0.0117, + "step": 94130 + }, + { + "epoch": 27.390165842304334, + "grad_norm": 0.15295293927192688, + "learning_rate": 3.6507482253948334e-06, + "loss": 0.0107, + "step": 94140 + }, + { + "epoch": 27.39307535641548, + "grad_norm": 0.10071603953838348, + "learning_rate": 3.6480458178179244e-06, + "loss": 0.0115, + "step": 94150 + }, + { + "epoch": 27.395984870526622, + "grad_norm": 0.13424327969551086, + "learning_rate": 3.6453442723470173e-06, + "loss": 0.0102, + "step": 94160 + }, + { + "epoch": 27.398894384637764, + "grad_norm": 0.09991291165351868, + "learning_rate": 3.6426435891872774e-06, + "loss": 0.0103, + "step": 94170 + }, + { + "epoch": 27.40180389874891, + "grad_norm": 0.08748897165060043, + "learning_rate": 3.639943768543809e-06, + "loss": 0.0121, + "step": 94180 + }, + { + "epoch": 27.40471341286005, + "grad_norm": 0.11095276474952698, + "learning_rate": 3.6372448106216373e-06, + "loss": 0.0107, + "step": 94190 + }, + { + "epoch": 27.407622926971197, + "grad_norm": 0.1247183158993721, + "learning_rate": 3.634546715625734e-06, + "loss": 0.0119, + "step": 94200 + }, + { + "epoch": 27.41053244108234, + "grad_norm": 0.11863432824611664, + "learning_rate": 3.631849483761007e-06, + "loss": 0.0122, + "step": 94210 + }, + { + "epoch": 27.413441955193484, + "grad_norm": 0.07958199828863144, + "learning_rate": 3.629153115232285e-06, + "loss": 0.0092, + "step": 94220 + }, + { + "epoch": 27.416351469304626, + "grad_norm": 0.08630666136741638, + "learning_rate": 3.6264576102443462e-06, + "loss": 0.011, + "step": 94230 + }, + { + "epoch": 27.419260983415768, + "grad_norm": 0.11470384895801544, + "learning_rate": 3.62376296900189e-06, + "loss": 0.0107, + "step": 94240 + }, + { + "epoch": 27.422170497526913, + "grad_norm": 0.06836704164743423, + "learning_rate": 3.6210691917095613e-06, + "loss": 0.0125, + "step": 94250 + }, + { + "epoch": 27.425080011638055, + "grad_norm": 0.1486886739730835, + "learning_rate": 3.6183762785719303e-06, + "loss": 0.0115, + "step": 94260 + }, + { + "epoch": 27.4279895257492, + "grad_norm": 0.10942312330007553, + "learning_rate": 3.615684229793514e-06, + "loss": 0.012, + "step": 94270 + }, + { + "epoch": 27.430899039860343, + "grad_norm": 0.1345135122537613, + "learning_rate": 3.6129930455787446e-06, + "loss": 0.0123, + "step": 94280 + }, + { + "epoch": 27.433808553971488, + "grad_norm": 0.11267916858196259, + "learning_rate": 3.6103027261320075e-06, + "loss": 0.012, + "step": 94290 + }, + { + "epoch": 27.43671806808263, + "grad_norm": 0.12139949947595596, + "learning_rate": 3.607613271657616e-06, + "loss": 0.011, + "step": 94300 + }, + { + "epoch": 27.439627582193772, + "grad_norm": 0.1378583014011383, + "learning_rate": 3.60492468235981e-06, + "loss": 0.0116, + "step": 94310 + }, + { + "epoch": 27.442537096304918, + "grad_norm": 0.13149873912334442, + "learning_rate": 3.6022369584427755e-06, + "loss": 0.013, + "step": 94320 + }, + { + "epoch": 27.44544661041606, + "grad_norm": 0.0947219654917717, + "learning_rate": 3.5995501001106234e-06, + "loss": 0.0111, + "step": 94330 + }, + { + "epoch": 27.448356124527205, + "grad_norm": 0.08003505319356918, + "learning_rate": 3.5968641075674068e-06, + "loss": 0.0103, + "step": 94340 + }, + { + "epoch": 27.451265638638347, + "grad_norm": 0.10896062850952148, + "learning_rate": 3.5941789810171043e-06, + "loss": 0.0102, + "step": 94350 + }, + { + "epoch": 27.454175152749492, + "grad_norm": 0.06824241578578949, + "learning_rate": 3.591494720663636e-06, + "loss": 0.0102, + "step": 94360 + }, + { + "epoch": 27.457084666860634, + "grad_norm": 0.08297097682952881, + "learning_rate": 3.588811326710859e-06, + "loss": 0.0104, + "step": 94370 + }, + { + "epoch": 27.459994180971776, + "grad_norm": 0.12354122847318649, + "learning_rate": 3.5861287993625536e-06, + "loss": 0.0112, + "step": 94380 + }, + { + "epoch": 27.46290369508292, + "grad_norm": 0.18749909102916718, + "learning_rate": 3.5834471388224334e-06, + "loss": 0.0122, + "step": 94390 + }, + { + "epoch": 27.465813209194064, + "grad_norm": 0.11915509402751923, + "learning_rate": 3.5807663452941685e-06, + "loss": 0.0114, + "step": 94400 + }, + { + "epoch": 27.46872272330521, + "grad_norm": 0.11411920189857483, + "learning_rate": 3.5780864189813382e-06, + "loss": 0.0092, + "step": 94410 + }, + { + "epoch": 27.47163223741635, + "grad_norm": 0.3275917172431946, + "learning_rate": 3.575407360087463e-06, + "loss": 0.0129, + "step": 94420 + }, + { + "epoch": 27.474541751527497, + "grad_norm": 0.08812204003334045, + "learning_rate": 3.572729168816002e-06, + "loss": 0.0108, + "step": 94430 + }, + { + "epoch": 27.47745126563864, + "grad_norm": 0.10042457282543182, + "learning_rate": 3.5700518453703533e-06, + "loss": 0.0128, + "step": 94440 + }, + { + "epoch": 27.48036077974978, + "grad_norm": 0.1212499663233757, + "learning_rate": 3.5673753899538352e-06, + "loss": 0.0116, + "step": 94450 + }, + { + "epoch": 27.483270293860926, + "grad_norm": 0.1081278845667839, + "learning_rate": 3.5646998027697012e-06, + "loss": 0.01, + "step": 94460 + }, + { + "epoch": 27.486179807972068, + "grad_norm": 0.09333217144012451, + "learning_rate": 3.562025084021153e-06, + "loss": 0.011, + "step": 94470 + }, + { + "epoch": 27.489089322083213, + "grad_norm": 0.11088661849498749, + "learning_rate": 3.55935123391132e-06, + "loss": 0.0126, + "step": 94480 + }, + { + "epoch": 27.491998836194355, + "grad_norm": 0.12610630691051483, + "learning_rate": 3.556678252643254e-06, + "loss": 0.0139, + "step": 94490 + }, + { + "epoch": 27.4949083503055, + "grad_norm": 0.137221559882164, + "learning_rate": 3.5540061404199543e-06, + "loss": 0.0117, + "step": 94500 + }, + { + "epoch": 27.497817864416643, + "grad_norm": 0.10144466906785965, + "learning_rate": 3.5513348974443565e-06, + "loss": 0.0108, + "step": 94510 + }, + { + "epoch": 27.500727378527785, + "grad_norm": 0.19792458415031433, + "learning_rate": 3.5486645239193206e-06, + "loss": 0.0109, + "step": 94520 + }, + { + "epoch": 27.50363689263893, + "grad_norm": 0.10946356505155563, + "learning_rate": 3.5459950200476356e-06, + "loss": 0.0106, + "step": 94530 + }, + { + "epoch": 27.506546406750072, + "grad_norm": 0.10513915866613388, + "learning_rate": 3.5433263860320393e-06, + "loss": 0.0112, + "step": 94540 + }, + { + "epoch": 27.509455920861217, + "grad_norm": 0.11364459246397018, + "learning_rate": 3.5406586220751995e-06, + "loss": 0.0105, + "step": 94550 + }, + { + "epoch": 27.51236543497236, + "grad_norm": 0.06953130662441254, + "learning_rate": 3.53799172837971e-06, + "loss": 0.01, + "step": 94560 + }, + { + "epoch": 27.5152749490835, + "grad_norm": 0.10619767010211945, + "learning_rate": 3.535325705148109e-06, + "loss": 0.0109, + "step": 94570 + }, + { + "epoch": 27.518184463194647, + "grad_norm": 0.11646179109811783, + "learning_rate": 3.5326605525828558e-06, + "loss": 0.0106, + "step": 94580 + }, + { + "epoch": 27.52109397730579, + "grad_norm": 0.08058073371648788, + "learning_rate": 3.5299962708863563e-06, + "loss": 0.0103, + "step": 94590 + }, + { + "epoch": 27.524003491416934, + "grad_norm": 0.09547862410545349, + "learning_rate": 3.5273328602609504e-06, + "loss": 0.0115, + "step": 94600 + }, + { + "epoch": 27.526913005528076, + "grad_norm": 0.09992432594299316, + "learning_rate": 3.524670320908895e-06, + "loss": 0.011, + "step": 94610 + }, + { + "epoch": 27.52982251963922, + "grad_norm": 0.12963585555553436, + "learning_rate": 3.5220086530324027e-06, + "loss": 0.0105, + "step": 94620 + }, + { + "epoch": 27.532732033750364, + "grad_norm": 0.1298833042383194, + "learning_rate": 3.5193478568336e-06, + "loss": 0.0125, + "step": 94630 + }, + { + "epoch": 27.535641547861506, + "grad_norm": 0.11170578002929688, + "learning_rate": 3.516687932514568e-06, + "loss": 0.0114, + "step": 94640 + }, + { + "epoch": 27.53855106197265, + "grad_norm": 0.1346655786037445, + "learning_rate": 3.5140288802772973e-06, + "loss": 0.0109, + "step": 94650 + }, + { + "epoch": 27.541460576083793, + "grad_norm": 0.08231014013290405, + "learning_rate": 3.5113707003237355e-06, + "loss": 0.0106, + "step": 94660 + }, + { + "epoch": 27.54437009019494, + "grad_norm": 0.10566341131925583, + "learning_rate": 3.5087133928557526e-06, + "loss": 0.0102, + "step": 94670 + }, + { + "epoch": 27.54727960430608, + "grad_norm": 0.12202590703964233, + "learning_rate": 3.506056958075152e-06, + "loss": 0.0108, + "step": 94680 + }, + { + "epoch": 27.550189118417226, + "grad_norm": 0.11222313344478607, + "learning_rate": 3.5034013961836687e-06, + "loss": 0.0138, + "step": 94690 + }, + { + "epoch": 27.553098632528368, + "grad_norm": 0.2648308277130127, + "learning_rate": 3.500746707382977e-06, + "loss": 0.0115, + "step": 94700 + }, + { + "epoch": 27.55600814663951, + "grad_norm": 0.11764099448919296, + "learning_rate": 3.49809289187469e-06, + "loss": 0.0107, + "step": 94710 + }, + { + "epoch": 27.558917660750655, + "grad_norm": 0.09712879359722137, + "learning_rate": 3.4954399498603373e-06, + "loss": 0.01, + "step": 94720 + }, + { + "epoch": 27.561827174861797, + "grad_norm": 0.08830895274877548, + "learning_rate": 3.4927878815413953e-06, + "loss": 0.0106, + "step": 94730 + }, + { + "epoch": 27.564736688972943, + "grad_norm": 0.10168620944023132, + "learning_rate": 3.4901366871192784e-06, + "loss": 0.0117, + "step": 94740 + }, + { + "epoch": 27.567646203084085, + "grad_norm": 0.09465060383081436, + "learning_rate": 3.4874863667953224e-06, + "loss": 0.0099, + "step": 94750 + }, + { + "epoch": 27.57055571719523, + "grad_norm": 0.13587167859077454, + "learning_rate": 3.4848369207707955e-06, + "loss": 0.0121, + "step": 94760 + }, + { + "epoch": 27.573465231306372, + "grad_norm": 0.1235727071762085, + "learning_rate": 3.482188349246912e-06, + "loss": 0.0126, + "step": 94770 + }, + { + "epoch": 27.576374745417514, + "grad_norm": 0.1250971257686615, + "learning_rate": 3.4795406524248162e-06, + "loss": 0.011, + "step": 94780 + }, + { + "epoch": 27.57928425952866, + "grad_norm": 0.11103999614715576, + "learning_rate": 3.476893830505579e-06, + "loss": 0.0111, + "step": 94790 + }, + { + "epoch": 27.5821937736398, + "grad_norm": 0.12277573347091675, + "learning_rate": 3.4742478836902013e-06, + "loss": 0.012, + "step": 94800 + }, + { + "epoch": 27.585103287750947, + "grad_norm": 0.1308079957962036, + "learning_rate": 3.4716028121796404e-06, + "loss": 0.0117, + "step": 94810 + }, + { + "epoch": 27.58801280186209, + "grad_norm": 0.08796503394842148, + "learning_rate": 3.468958616174768e-06, + "loss": 0.0098, + "step": 94820 + }, + { + "epoch": 27.59092231597323, + "grad_norm": 0.10759902000427246, + "learning_rate": 3.4663152958763838e-06, + "loss": 0.0094, + "step": 94830 + }, + { + "epoch": 27.593831830084376, + "grad_norm": 0.15412397682666779, + "learning_rate": 3.4636728514852383e-06, + "loss": 0.0099, + "step": 94840 + }, + { + "epoch": 27.596741344195518, + "grad_norm": 0.13122279942035675, + "learning_rate": 3.4610312832020106e-06, + "loss": 0.01, + "step": 94850 + }, + { + "epoch": 27.599650858306664, + "grad_norm": 0.09805316478013992, + "learning_rate": 3.458390591227302e-06, + "loss": 0.0102, + "step": 94860 + }, + { + "epoch": 27.602560372417805, + "grad_norm": 0.07100105285644531, + "learning_rate": 3.455750775761666e-06, + "loss": 0.0113, + "step": 94870 + }, + { + "epoch": 27.60546988652895, + "grad_norm": 0.10157051682472229, + "learning_rate": 3.453111837005566e-06, + "loss": 0.0126, + "step": 94880 + }, + { + "epoch": 27.608379400640093, + "grad_norm": 0.10816706717014313, + "learning_rate": 3.4504737751594257e-06, + "loss": 0.0106, + "step": 94890 + }, + { + "epoch": 27.611288914751235, + "grad_norm": 0.11556370556354523, + "learning_rate": 3.4478365904235764e-06, + "loss": 0.0116, + "step": 94900 + }, + { + "epoch": 27.61419842886238, + "grad_norm": 0.17917726933956146, + "learning_rate": 3.4452002829983016e-06, + "loss": 0.0129, + "step": 94910 + }, + { + "epoch": 27.617107942973522, + "grad_norm": 0.06500552594661713, + "learning_rate": 3.4425648530838124e-06, + "loss": 0.0092, + "step": 94920 + }, + { + "epoch": 27.620017457084668, + "grad_norm": 0.10232625156641006, + "learning_rate": 3.4399303008802486e-06, + "loss": 0.0114, + "step": 94930 + }, + { + "epoch": 27.62292697119581, + "grad_norm": 0.08868725597858429, + "learning_rate": 3.4372966265876905e-06, + "loss": 0.0106, + "step": 94940 + }, + { + "epoch": 27.625836485306955, + "grad_norm": 0.12720325589179993, + "learning_rate": 3.4346638304061428e-06, + "loss": 0.0117, + "step": 94950 + }, + { + "epoch": 27.628745999418097, + "grad_norm": 0.07546389847993851, + "learning_rate": 3.4320319125355544e-06, + "loss": 0.0104, + "step": 94960 + }, + { + "epoch": 27.63165551352924, + "grad_norm": 0.10670248419046402, + "learning_rate": 3.429400873175803e-06, + "loss": 0.0119, + "step": 94970 + }, + { + "epoch": 27.634565027640384, + "grad_norm": 0.11871220916509628, + "learning_rate": 3.426770712526696e-06, + "loss": 0.0115, + "step": 94980 + }, + { + "epoch": 27.637474541751526, + "grad_norm": 0.1486886739730835, + "learning_rate": 3.424141430787972e-06, + "loss": 0.0119, + "step": 94990 + }, + { + "epoch": 27.640384055862672, + "grad_norm": 0.10083924233913422, + "learning_rate": 3.4215130281593125e-06, + "loss": 0.0115, + "step": 95000 + }, + { + "epoch": 27.643293569973814, + "grad_norm": 0.10269910842180252, + "learning_rate": 3.4188855048403316e-06, + "loss": 0.0104, + "step": 95010 + }, + { + "epoch": 27.64620308408496, + "grad_norm": 0.11101341992616653, + "learning_rate": 3.4162588610305643e-06, + "loss": 0.0086, + "step": 95020 + }, + { + "epoch": 27.6491125981961, + "grad_norm": 0.07598558068275452, + "learning_rate": 3.41363309692949e-06, + "loss": 0.0131, + "step": 95030 + }, + { + "epoch": 27.652022112307243, + "grad_norm": 0.1205715462565422, + "learning_rate": 3.4110082127365233e-06, + "loss": 0.0126, + "step": 95040 + }, + { + "epoch": 27.65493162641839, + "grad_norm": 0.12604796886444092, + "learning_rate": 3.408384208651002e-06, + "loss": 0.0103, + "step": 95050 + }, + { + "epoch": 27.65784114052953, + "grad_norm": 0.11156399548053741, + "learning_rate": 3.4057610848721965e-06, + "loss": 0.0115, + "step": 95060 + }, + { + "epoch": 27.660750654640676, + "grad_norm": 0.08702186495065689, + "learning_rate": 3.4031388415993227e-06, + "loss": 0.0104, + "step": 95070 + }, + { + "epoch": 27.663660168751818, + "grad_norm": 0.06884746998548508, + "learning_rate": 3.400517479031524e-06, + "loss": 0.0123, + "step": 95080 + }, + { + "epoch": 27.666569682862963, + "grad_norm": 0.13732360303401947, + "learning_rate": 3.3978969973678754e-06, + "loss": 0.0111, + "step": 95090 + }, + { + "epoch": 27.669479196974105, + "grad_norm": 0.12410221248865128, + "learning_rate": 3.395277396807372e-06, + "loss": 0.0135, + "step": 95100 + }, + { + "epoch": 27.672388711085247, + "grad_norm": 0.11395318806171417, + "learning_rate": 3.392658677548975e-06, + "loss": 0.0118, + "step": 95110 + }, + { + "epoch": 27.675298225196393, + "grad_norm": 0.16077199578285217, + "learning_rate": 3.39004083979155e-06, + "loss": 0.01, + "step": 95120 + }, + { + "epoch": 27.678207739307535, + "grad_norm": 0.11561659723520279, + "learning_rate": 3.3874238837339e-06, + "loss": 0.0107, + "step": 95130 + }, + { + "epoch": 27.68111725341868, + "grad_norm": 0.09501642733812332, + "learning_rate": 3.384807809574769e-06, + "loss": 0.01, + "step": 95140 + }, + { + "epoch": 27.684026767529822, + "grad_norm": 0.11177954077720642, + "learning_rate": 3.3821926175128355e-06, + "loss": 0.0123, + "step": 95150 + }, + { + "epoch": 27.686936281640968, + "grad_norm": 0.07810449600219727, + "learning_rate": 3.3795783077467015e-06, + "loss": 0.0117, + "step": 95160 + }, + { + "epoch": 27.68984579575211, + "grad_norm": 0.1867227703332901, + "learning_rate": 3.376964880474903e-06, + "loss": 0.0121, + "step": 95170 + }, + { + "epoch": 27.69275530986325, + "grad_norm": 0.09648644924163818, + "learning_rate": 3.374352335895917e-06, + "loss": 0.0102, + "step": 95180 + }, + { + "epoch": 27.695664823974397, + "grad_norm": 0.11608356237411499, + "learning_rate": 3.371740674208153e-06, + "loss": 0.01, + "step": 95190 + }, + { + "epoch": 27.69857433808554, + "grad_norm": 0.09437508136034012, + "learning_rate": 3.36912989560994e-06, + "loss": 0.012, + "step": 95200 + }, + { + "epoch": 27.701483852196684, + "grad_norm": 0.12725184857845306, + "learning_rate": 3.366520000299554e-06, + "loss": 0.0109, + "step": 95210 + }, + { + "epoch": 27.704393366307826, + "grad_norm": 0.11633815616369247, + "learning_rate": 3.363910988475205e-06, + "loss": 0.0108, + "step": 95220 + }, + { + "epoch": 27.707302880418972, + "grad_norm": 0.06640446186065674, + "learning_rate": 3.3613028603350203e-06, + "loss": 0.0122, + "step": 95230 + }, + { + "epoch": 27.710212394530114, + "grad_norm": 0.11604632437229156, + "learning_rate": 3.358695616077078e-06, + "loss": 0.011, + "step": 95240 + }, + { + "epoch": 27.713121908641256, + "grad_norm": 0.12904050946235657, + "learning_rate": 3.356089255899375e-06, + "loss": 0.0124, + "step": 95250 + }, + { + "epoch": 27.7160314227524, + "grad_norm": 0.11814986914396286, + "learning_rate": 3.3534837799998524e-06, + "loss": 0.0103, + "step": 95260 + }, + { + "epoch": 27.718940936863543, + "grad_norm": 0.09813809394836426, + "learning_rate": 3.350879188576374e-06, + "loss": 0.0118, + "step": 95270 + }, + { + "epoch": 27.72185045097469, + "grad_norm": 0.10894063115119934, + "learning_rate": 3.348275481826746e-06, + "loss": 0.0116, + "step": 95280 + }, + { + "epoch": 27.72475996508583, + "grad_norm": 0.09469447284936905, + "learning_rate": 3.345672659948698e-06, + "loss": 0.0109, + "step": 95290 + }, + { + "epoch": 27.727669479196972, + "grad_norm": 0.14661753177642822, + "learning_rate": 3.3430707231398994e-06, + "loss": 0.0148, + "step": 95300 + }, + { + "epoch": 27.730578993308118, + "grad_norm": 0.1303730309009552, + "learning_rate": 3.340469671597954e-06, + "loss": 0.0106, + "step": 95310 + }, + { + "epoch": 27.73348850741926, + "grad_norm": 0.09840922057628632, + "learning_rate": 3.3378695055203866e-06, + "loss": 0.0108, + "step": 95320 + }, + { + "epoch": 27.736398021530405, + "grad_norm": 0.0980444848537445, + "learning_rate": 3.335270225104667e-06, + "loss": 0.0109, + "step": 95330 + }, + { + "epoch": 27.739307535641547, + "grad_norm": 0.13783204555511475, + "learning_rate": 3.332671830548198e-06, + "loss": 0.0118, + "step": 95340 + }, + { + "epoch": 27.742217049752693, + "grad_norm": 0.14507131278514862, + "learning_rate": 3.3300743220483037e-06, + "loss": 0.0124, + "step": 95350 + }, + { + "epoch": 27.745126563863835, + "grad_norm": 0.08336365967988968, + "learning_rate": 3.327477699802248e-06, + "loss": 0.0093, + "step": 95360 + }, + { + "epoch": 27.748036077974977, + "grad_norm": 0.13848307728767395, + "learning_rate": 3.3248819640072285e-06, + "loss": 0.0106, + "step": 95370 + }, + { + "epoch": 27.750945592086122, + "grad_norm": 0.16974091529846191, + "learning_rate": 3.322287114860379e-06, + "loss": 0.0133, + "step": 95380 + }, + { + "epoch": 27.753855106197264, + "grad_norm": 0.09915503114461899, + "learning_rate": 3.3196931525587566e-06, + "loss": 0.0101, + "step": 95390 + }, + { + "epoch": 27.75676462030841, + "grad_norm": 0.12499987334012985, + "learning_rate": 3.317100077299347e-06, + "loss": 0.0101, + "step": 95400 + }, + { + "epoch": 27.75967413441955, + "grad_norm": 0.08609102666378021, + "learning_rate": 3.3145078892790947e-06, + "loss": 0.0114, + "step": 95410 + }, + { + "epoch": 27.762583648530697, + "grad_norm": 0.14108934998512268, + "learning_rate": 3.311916588694852e-06, + "loss": 0.0103, + "step": 95420 + }, + { + "epoch": 27.76549316264184, + "grad_norm": 0.11490854620933533, + "learning_rate": 3.3093261757434035e-06, + "loss": 0.0099, + "step": 95430 + }, + { + "epoch": 27.76840267675298, + "grad_norm": 0.09430652856826782, + "learning_rate": 3.3067366506214814e-06, + "loss": 0.0107, + "step": 95440 + }, + { + "epoch": 27.771312190864126, + "grad_norm": 0.08889011293649673, + "learning_rate": 3.304148013525747e-06, + "loss": 0.0114, + "step": 95450 + }, + { + "epoch": 27.774221704975268, + "grad_norm": 0.09472872316837311, + "learning_rate": 3.3015602646527843e-06, + "loss": 0.0109, + "step": 95460 + }, + { + "epoch": 27.777131219086414, + "grad_norm": 0.06845200061798096, + "learning_rate": 3.2989734041991114e-06, + "loss": 0.0108, + "step": 95470 + }, + { + "epoch": 27.780040733197556, + "grad_norm": 0.14440305531024933, + "learning_rate": 3.2963874323611897e-06, + "loss": 0.0111, + "step": 95480 + }, + { + "epoch": 27.7829502473087, + "grad_norm": 0.13902699947357178, + "learning_rate": 3.2938023493354086e-06, + "loss": 0.0121, + "step": 95490 + }, + { + "epoch": 27.785859761419843, + "grad_norm": 0.06112741306424141, + "learning_rate": 3.291218155318081e-06, + "loss": 0.0097, + "step": 95500 + }, + { + "epoch": 27.788769275530985, + "grad_norm": 0.10871411859989166, + "learning_rate": 3.2886348505054633e-06, + "loss": 0.0114, + "step": 95510 + }, + { + "epoch": 27.79167878964213, + "grad_norm": 0.1257927268743515, + "learning_rate": 3.286052435093745e-06, + "loss": 0.0117, + "step": 95520 + }, + { + "epoch": 27.794588303753272, + "grad_norm": 0.15815994143486023, + "learning_rate": 3.283470909279038e-06, + "loss": 0.0129, + "step": 95530 + }, + { + "epoch": 27.797497817864418, + "grad_norm": 0.0984252318739891, + "learning_rate": 3.2808902732573896e-06, + "loss": 0.0131, + "step": 95540 + }, + { + "epoch": 27.80040733197556, + "grad_norm": 0.10689353942871094, + "learning_rate": 3.278310527224784e-06, + "loss": 0.0114, + "step": 95550 + }, + { + "epoch": 27.8033168460867, + "grad_norm": 0.05862057954072952, + "learning_rate": 3.2757316713771444e-06, + "loss": 0.0095, + "step": 95560 + }, + { + "epoch": 27.806226360197847, + "grad_norm": 0.1461976319551468, + "learning_rate": 3.2731537059103054e-06, + "loss": 0.0092, + "step": 95570 + }, + { + "epoch": 27.80913587430899, + "grad_norm": 0.10165470093488693, + "learning_rate": 3.2705766310200557e-06, + "loss": 0.0111, + "step": 95580 + }, + { + "epoch": 27.812045388420135, + "grad_norm": 0.20507851243019104, + "learning_rate": 3.2680004469021003e-06, + "loss": 0.0102, + "step": 95590 + }, + { + "epoch": 27.814954902531277, + "grad_norm": 0.08794418722391129, + "learning_rate": 3.265425153752086e-06, + "loss": 0.0114, + "step": 95600 + }, + { + "epoch": 27.817864416642422, + "grad_norm": 0.09608937054872513, + "learning_rate": 3.2628507517655927e-06, + "loss": 0.0115, + "step": 95610 + }, + { + "epoch": 27.820773930753564, + "grad_norm": 0.07978928089141846, + "learning_rate": 3.2602772411381243e-06, + "loss": 0.0132, + "step": 95620 + }, + { + "epoch": 27.823683444864706, + "grad_norm": 0.08332649618387222, + "learning_rate": 3.2577046220651264e-06, + "loss": 0.0113, + "step": 95630 + }, + { + "epoch": 27.82659295897585, + "grad_norm": 0.08877512067556381, + "learning_rate": 3.2551328947419675e-06, + "loss": 0.0114, + "step": 95640 + }, + { + "epoch": 27.829502473086993, + "grad_norm": 0.13399377465248108, + "learning_rate": 3.252562059363959e-06, + "loss": 0.0134, + "step": 95650 + }, + { + "epoch": 27.83241198719814, + "grad_norm": 0.14602626860141754, + "learning_rate": 3.249992116126331e-06, + "loss": 0.0126, + "step": 95660 + }, + { + "epoch": 27.83532150130928, + "grad_norm": 0.09860152006149292, + "learning_rate": 3.24742306522426e-06, + "loss": 0.0114, + "step": 95670 + }, + { + "epoch": 27.838231015420426, + "grad_norm": 0.07603809982538223, + "learning_rate": 3.2448549068528505e-06, + "loss": 0.0117, + "step": 95680 + }, + { + "epoch": 27.841140529531568, + "grad_norm": 0.14143702387809753, + "learning_rate": 3.2422876412071324e-06, + "loss": 0.0122, + "step": 95690 + }, + { + "epoch": 27.84405004364271, + "grad_norm": 0.08508665859699249, + "learning_rate": 3.2397212684820665e-06, + "loss": 0.0108, + "step": 95700 + }, + { + "epoch": 27.846959557753856, + "grad_norm": 0.3370261788368225, + "learning_rate": 3.237155788872569e-06, + "loss": 0.0098, + "step": 95710 + }, + { + "epoch": 27.849869071864997, + "grad_norm": 0.13603205978870392, + "learning_rate": 3.2345912025734586e-06, + "loss": 0.0112, + "step": 95720 + }, + { + "epoch": 27.852778585976143, + "grad_norm": 0.10424172878265381, + "learning_rate": 3.2320275097794997e-06, + "loss": 0.0118, + "step": 95730 + }, + { + "epoch": 27.855688100087285, + "grad_norm": 0.09976571053266525, + "learning_rate": 3.229464710685389e-06, + "loss": 0.0101, + "step": 95740 + }, + { + "epoch": 27.85859761419843, + "grad_norm": 0.14630518853664398, + "learning_rate": 3.2269028054857576e-06, + "loss": 0.0117, + "step": 95750 + }, + { + "epoch": 27.861507128309572, + "grad_norm": 0.08400361984968185, + "learning_rate": 3.2243417943751623e-06, + "loss": 0.011, + "step": 95760 + }, + { + "epoch": 27.864416642420714, + "grad_norm": 0.10217839479446411, + "learning_rate": 3.221781677548093e-06, + "loss": 0.0117, + "step": 95770 + }, + { + "epoch": 27.86732615653186, + "grad_norm": 0.1014319434762001, + "learning_rate": 3.2192224551989734e-06, + "loss": 0.011, + "step": 95780 + }, + { + "epoch": 27.870235670643, + "grad_norm": 0.10091187059879303, + "learning_rate": 3.2166641275221688e-06, + "loss": 0.012, + "step": 95790 + }, + { + "epoch": 27.873145184754147, + "grad_norm": 0.074745312333107, + "learning_rate": 3.2141066947119547e-06, + "loss": 0.0105, + "step": 95800 + }, + { + "epoch": 27.87605469886529, + "grad_norm": 0.11889063566923141, + "learning_rate": 3.211550156962556e-06, + "loss": 0.011, + "step": 95810 + }, + { + "epoch": 27.878964212976435, + "grad_norm": 0.12420188635587692, + "learning_rate": 3.208994514468131e-06, + "loss": 0.0095, + "step": 95820 + }, + { + "epoch": 27.881873727087576, + "grad_norm": 0.09920429438352585, + "learning_rate": 3.206439767422759e-06, + "loss": 0.0116, + "step": 95830 + }, + { + "epoch": 27.88478324119872, + "grad_norm": 0.1052580177783966, + "learning_rate": 3.203885916020452e-06, + "loss": 0.0122, + "step": 95840 + }, + { + "epoch": 27.887692755309864, + "grad_norm": 0.112399160861969, + "learning_rate": 3.2013329604551628e-06, + "loss": 0.0133, + "step": 95850 + }, + { + "epoch": 27.890602269421006, + "grad_norm": 0.0980461910367012, + "learning_rate": 3.198780900920773e-06, + "loss": 0.0132, + "step": 95860 + }, + { + "epoch": 27.89351178353215, + "grad_norm": 0.08061207085847855, + "learning_rate": 3.1962297376110896e-06, + "loss": 0.0125, + "step": 95870 + }, + { + "epoch": 27.896421297643293, + "grad_norm": 0.07072219997644424, + "learning_rate": 3.1936794707198635e-06, + "loss": 0.0095, + "step": 95880 + }, + { + "epoch": 27.89933081175444, + "grad_norm": 0.09454259276390076, + "learning_rate": 3.1911301004407622e-06, + "loss": 0.0108, + "step": 95890 + }, + { + "epoch": 27.90224032586558, + "grad_norm": 0.07414795458316803, + "learning_rate": 3.188581626967402e-06, + "loss": 0.0121, + "step": 95900 + }, + { + "epoch": 27.905149839976723, + "grad_norm": 0.11219244450330734, + "learning_rate": 3.1860340504933155e-06, + "loss": 0.0096, + "step": 95910 + }, + { + "epoch": 27.908059354087868, + "grad_norm": 0.1459604650735855, + "learning_rate": 3.1834873712119773e-06, + "loss": 0.0114, + "step": 95920 + }, + { + "epoch": 27.91096886819901, + "grad_norm": 0.09956956654787064, + "learning_rate": 3.1809415893167944e-06, + "loss": 0.0108, + "step": 95930 + }, + { + "epoch": 27.913878382310156, + "grad_norm": 0.12410473823547363, + "learning_rate": 3.1783967050010954e-06, + "loss": 0.012, + "step": 95940 + }, + { + "epoch": 27.916787896421297, + "grad_norm": 0.09649443626403809, + "learning_rate": 3.1758527184581544e-06, + "loss": 0.0121, + "step": 95950 + }, + { + "epoch": 27.91969741053244, + "grad_norm": 0.10246909409761429, + "learning_rate": 3.173309629881163e-06, + "loss": 0.0131, + "step": 95960 + }, + { + "epoch": 27.922606924643585, + "grad_norm": 0.10874198377132416, + "learning_rate": 3.170767439463256e-06, + "loss": 0.0099, + "step": 95970 + }, + { + "epoch": 27.925516438754727, + "grad_norm": 0.10506575554609299, + "learning_rate": 3.1682261473974995e-06, + "loss": 0.0118, + "step": 95980 + }, + { + "epoch": 27.928425952865872, + "grad_norm": 0.09434299916028976, + "learning_rate": 3.165685753876885e-06, + "loss": 0.0117, + "step": 95990 + }, + { + "epoch": 27.931335466977014, + "grad_norm": 0.15947142243385315, + "learning_rate": 3.1631462590943337e-06, + "loss": 0.011, + "step": 96000 + }, + { + "epoch": 27.93424498108816, + "grad_norm": 0.12598560750484467, + "learning_rate": 3.160607663242706e-06, + "loss": 0.0101, + "step": 96010 + }, + { + "epoch": 27.9371544951993, + "grad_norm": 0.0969325453042984, + "learning_rate": 3.1580699665147987e-06, + "loss": 0.0121, + "step": 96020 + }, + { + "epoch": 27.940064009310444, + "grad_norm": 0.09397216886281967, + "learning_rate": 3.155533169103323e-06, + "loss": 0.011, + "step": 96030 + }, + { + "epoch": 27.94297352342159, + "grad_norm": 0.12951764464378357, + "learning_rate": 3.1529972712009358e-06, + "loss": 0.0094, + "step": 96040 + }, + { + "epoch": 27.94588303753273, + "grad_norm": 0.10417152941226959, + "learning_rate": 3.1504622730002287e-06, + "loss": 0.0108, + "step": 96050 + }, + { + "epoch": 27.948792551643876, + "grad_norm": 0.09746059030294418, + "learning_rate": 3.147928174693709e-06, + "loss": 0.0113, + "step": 96060 + }, + { + "epoch": 27.95170206575502, + "grad_norm": 0.07501856982707977, + "learning_rate": 3.145394976473826e-06, + "loss": 0.0117, + "step": 96070 + }, + { + "epoch": 27.954611579866164, + "grad_norm": 0.16481079161167145, + "learning_rate": 3.1428626785329594e-06, + "loss": 0.0096, + "step": 96080 + }, + { + "epoch": 27.957521093977306, + "grad_norm": 0.22736619412899017, + "learning_rate": 3.1403312810634255e-06, + "loss": 0.0108, + "step": 96090 + }, + { + "epoch": 27.960430608088448, + "grad_norm": 0.10124709457159042, + "learning_rate": 3.137800784257465e-06, + "loss": 0.0111, + "step": 96100 + }, + { + "epoch": 27.963340122199593, + "grad_norm": 0.1162559986114502, + "learning_rate": 3.1352711883072434e-06, + "loss": 0.0099, + "step": 96110 + }, + { + "epoch": 27.966249636310735, + "grad_norm": 0.08285607397556305, + "learning_rate": 3.132742493404881e-06, + "loss": 0.0093, + "step": 96120 + }, + { + "epoch": 27.96915915042188, + "grad_norm": 0.23184438049793243, + "learning_rate": 3.1302146997424114e-06, + "loss": 0.0102, + "step": 96130 + }, + { + "epoch": 27.972068664533023, + "grad_norm": 0.15531349182128906, + "learning_rate": 3.127687807511796e-06, + "loss": 0.0092, + "step": 96140 + }, + { + "epoch": 27.974978178644168, + "grad_norm": 0.09367205202579498, + "learning_rate": 3.1251618169049413e-06, + "loss": 0.0091, + "step": 96150 + }, + { + "epoch": 27.97788769275531, + "grad_norm": 0.0954299196600914, + "learning_rate": 3.1226367281136835e-06, + "loss": 0.0105, + "step": 96160 + }, + { + "epoch": 27.980797206866452, + "grad_norm": 0.08079216629266739, + "learning_rate": 3.12011254132978e-06, + "loss": 0.0107, + "step": 96170 + }, + { + "epoch": 27.983706720977597, + "grad_norm": 0.13292697072029114, + "learning_rate": 3.1175892567449316e-06, + "loss": 0.0107, + "step": 96180 + }, + { + "epoch": 27.98661623508874, + "grad_norm": 0.11988982558250427, + "learning_rate": 3.1150668745507573e-06, + "loss": 0.0104, + "step": 96190 + }, + { + "epoch": 27.989525749199885, + "grad_norm": 0.0814858004450798, + "learning_rate": 3.112545394938825e-06, + "loss": 0.0106, + "step": 96200 + }, + { + "epoch": 27.992435263311027, + "grad_norm": 0.20008736848831177, + "learning_rate": 3.110024818100615e-06, + "loss": 0.0106, + "step": 96210 + }, + { + "epoch": 27.995344777422172, + "grad_norm": 0.10253681242465973, + "learning_rate": 3.107505144227553e-06, + "loss": 0.0119, + "step": 96220 + }, + { + "epoch": 27.998254291533314, + "grad_norm": 0.12773674726486206, + "learning_rate": 3.104986373510996e-06, + "loss": 0.0095, + "step": 96230 + }, + { + "epoch": 28.001163805644456, + "grad_norm": 0.15294595062732697, + "learning_rate": 3.1024685061422196e-06, + "loss": 0.0093, + "step": 96240 + }, + { + "epoch": 28.0040733197556, + "grad_norm": 0.1709751933813095, + "learning_rate": 3.0999515423124473e-06, + "loss": 0.0118, + "step": 96250 + }, + { + "epoch": 28.006982833866743, + "grad_norm": 0.18763971328735352, + "learning_rate": 3.0974354822128174e-06, + "loss": 0.011, + "step": 96260 + }, + { + "epoch": 28.00989234797789, + "grad_norm": 0.0964696928858757, + "learning_rate": 3.094920326034417e-06, + "loss": 0.011, + "step": 96270 + }, + { + "epoch": 28.01280186208903, + "grad_norm": 0.09700702130794525, + "learning_rate": 3.092406073968247e-06, + "loss": 0.0123, + "step": 96280 + }, + { + "epoch": 28.015711376200173, + "grad_norm": 0.08651600778102875, + "learning_rate": 3.089892726205257e-06, + "loss": 0.0101, + "step": 96290 + }, + { + "epoch": 28.01862089031132, + "grad_norm": 0.05795402079820633, + "learning_rate": 3.0873802829363108e-06, + "loss": 0.0107, + "step": 96300 + }, + { + "epoch": 28.02153040442246, + "grad_norm": 0.1959853172302246, + "learning_rate": 3.0848687443522145e-06, + "loss": 0.01, + "step": 96310 + }, + { + "epoch": 28.024439918533606, + "grad_norm": 0.10285027325153351, + "learning_rate": 3.0823581106437093e-06, + "loss": 0.0111, + "step": 96320 + }, + { + "epoch": 28.027349432644748, + "grad_norm": 0.10721846669912338, + "learning_rate": 3.079848382001451e-06, + "loss": 0.011, + "step": 96330 + }, + { + "epoch": 28.030258946755893, + "grad_norm": 0.09539555013179779, + "learning_rate": 3.0773395586160423e-06, + "loss": 0.0102, + "step": 96340 + }, + { + "epoch": 28.033168460867035, + "grad_norm": 0.09827952086925507, + "learning_rate": 3.074831640678014e-06, + "loss": 0.0088, + "step": 96350 + }, + { + "epoch": 28.036077974978177, + "grad_norm": 0.09395992755889893, + "learning_rate": 3.0723246283778256e-06, + "loss": 0.0102, + "step": 96360 + }, + { + "epoch": 28.038987489089322, + "grad_norm": 0.11089697480201721, + "learning_rate": 3.069818521905861e-06, + "loss": 0.0122, + "step": 96370 + }, + { + "epoch": 28.041897003200464, + "grad_norm": 0.06394951790571213, + "learning_rate": 3.0673133214524467e-06, + "loss": 0.0122, + "step": 96380 + }, + { + "epoch": 28.04480651731161, + "grad_norm": 0.10737595707178116, + "learning_rate": 3.0648090272078404e-06, + "loss": 0.0109, + "step": 96390 + }, + { + "epoch": 28.047716031422752, + "grad_norm": 0.09763885289430618, + "learning_rate": 3.062305639362225e-06, + "loss": 0.0091, + "step": 96400 + }, + { + "epoch": 28.050625545533897, + "grad_norm": 0.12139733880758286, + "learning_rate": 3.059803158105705e-06, + "loss": 0.0122, + "step": 96410 + }, + { + "epoch": 28.05353505964504, + "grad_norm": 0.14718031883239746, + "learning_rate": 3.0573015836283454e-06, + "loss": 0.0123, + "step": 96420 + }, + { + "epoch": 28.05644457375618, + "grad_norm": 0.13033831119537354, + "learning_rate": 3.054800916120115e-06, + "loss": 0.0115, + "step": 96430 + }, + { + "epoch": 28.059354087867327, + "grad_norm": 0.07716056704521179, + "learning_rate": 3.0523011557709195e-06, + "loss": 0.0121, + "step": 96440 + }, + { + "epoch": 28.06226360197847, + "grad_norm": 0.12140464782714844, + "learning_rate": 3.0498023027706036e-06, + "loss": 0.0123, + "step": 96450 + }, + { + "epoch": 28.065173116089614, + "grad_norm": 0.13308554887771606, + "learning_rate": 3.0473043573089414e-06, + "loss": 0.0118, + "step": 96460 + }, + { + "epoch": 28.068082630200756, + "grad_norm": 0.1049046739935875, + "learning_rate": 3.044807319575634e-06, + "loss": 0.0114, + "step": 96470 + }, + { + "epoch": 28.0709921443119, + "grad_norm": 0.0995122641324997, + "learning_rate": 3.0423111897603085e-06, + "loss": 0.0116, + "step": 96480 + }, + { + "epoch": 28.073901658423043, + "grad_norm": 0.08768175542354584, + "learning_rate": 3.039815968052534e-06, + "loss": 0.0093, + "step": 96490 + }, + { + "epoch": 28.076811172534185, + "grad_norm": 0.0760047435760498, + "learning_rate": 3.0373216546418105e-06, + "loss": 0.0101, + "step": 96500 + }, + { + "epoch": 28.07972068664533, + "grad_norm": 0.09915129840373993, + "learning_rate": 3.034828249717556e-06, + "loss": 0.0125, + "step": 96510 + }, + { + "epoch": 28.082630200756473, + "grad_norm": 0.1715427041053772, + "learning_rate": 3.032335753469135e-06, + "loss": 0.0094, + "step": 96520 + }, + { + "epoch": 28.08553971486762, + "grad_norm": 0.13833710551261902, + "learning_rate": 3.029844166085835e-06, + "loss": 0.0109, + "step": 96530 + }, + { + "epoch": 28.08844922897876, + "grad_norm": 0.07223451137542725, + "learning_rate": 3.027353487756872e-06, + "loss": 0.0115, + "step": 96540 + }, + { + "epoch": 28.091358743089906, + "grad_norm": 0.10387324541807175, + "learning_rate": 3.0248637186714027e-06, + "loss": 0.0117, + "step": 96550 + }, + { + "epoch": 28.094268257201048, + "grad_norm": 0.12927110493183136, + "learning_rate": 3.0223748590184997e-06, + "loss": 0.0108, + "step": 96560 + }, + { + "epoch": 28.09717777131219, + "grad_norm": 0.10737679898738861, + "learning_rate": 3.019886908987186e-06, + "loss": 0.0098, + "step": 96570 + }, + { + "epoch": 28.100087285423335, + "grad_norm": 0.09903975576162338, + "learning_rate": 3.0173998687663963e-06, + "loss": 0.012, + "step": 96580 + }, + { + "epoch": 28.102996799534477, + "grad_norm": 0.09906672686338425, + "learning_rate": 3.0149137385450104e-06, + "loss": 0.0101, + "step": 96590 + }, + { + "epoch": 28.105906313645622, + "grad_norm": 0.1351672261953354, + "learning_rate": 3.0124285185118295e-06, + "loss": 0.0115, + "step": 96600 + }, + { + "epoch": 28.108815827756764, + "grad_norm": 0.1094542145729065, + "learning_rate": 3.0099442088555906e-06, + "loss": 0.0113, + "step": 96610 + }, + { + "epoch": 28.11172534186791, + "grad_norm": 0.07726401090621948, + "learning_rate": 3.007460809764965e-06, + "loss": 0.0123, + "step": 96620 + }, + { + "epoch": 28.11463485597905, + "grad_norm": 0.14052297174930573, + "learning_rate": 3.004978321428544e-06, + "loss": 0.0115, + "step": 96630 + }, + { + "epoch": 28.117544370090194, + "grad_norm": 0.13830910623073578, + "learning_rate": 3.0024967440348585e-06, + "loss": 0.0145, + "step": 96640 + }, + { + "epoch": 28.12045388420134, + "grad_norm": 0.22034676373004913, + "learning_rate": 3.0000160777723728e-06, + "loss": 0.0096, + "step": 96650 + }, + { + "epoch": 28.12336339831248, + "grad_norm": 0.08394941687583923, + "learning_rate": 2.997536322829472e-06, + "loss": 0.0116, + "step": 96660 + }, + { + "epoch": 28.126272912423627, + "grad_norm": 0.07819589972496033, + "learning_rate": 2.995057479394476e-06, + "loss": 0.0088, + "step": 96670 + }, + { + "epoch": 28.12918242653477, + "grad_norm": 0.12791401147842407, + "learning_rate": 2.9925795476556378e-06, + "loss": 0.013, + "step": 96680 + }, + { + "epoch": 28.13209194064591, + "grad_norm": 0.16163764894008636, + "learning_rate": 2.9901025278011444e-06, + "loss": 0.0107, + "step": 96690 + }, + { + "epoch": 28.135001454757056, + "grad_norm": 0.1316872537136078, + "learning_rate": 2.987626420019107e-06, + "loss": 0.0107, + "step": 96700 + }, + { + "epoch": 28.137910968868198, + "grad_norm": 0.5938172340393066, + "learning_rate": 2.985151224497559e-06, + "loss": 0.0125, + "step": 96710 + }, + { + "epoch": 28.140820482979343, + "grad_norm": 0.18991714715957642, + "learning_rate": 2.982676941424492e-06, + "loss": 0.0127, + "step": 96720 + }, + { + "epoch": 28.143729997090485, + "grad_norm": 0.07440955936908722, + "learning_rate": 2.980203570987805e-06, + "loss": 0.0118, + "step": 96730 + }, + { + "epoch": 28.14663951120163, + "grad_norm": 0.09816870093345642, + "learning_rate": 2.977731113375329e-06, + "loss": 0.0106, + "step": 96740 + }, + { + "epoch": 28.149549025312773, + "grad_norm": 0.10693905502557755, + "learning_rate": 2.975259568774835e-06, + "loss": 0.0108, + "step": 96750 + }, + { + "epoch": 28.152458539423915, + "grad_norm": 0.20695947110652924, + "learning_rate": 2.972788937374024e-06, + "loss": 0.0122, + "step": 96760 + }, + { + "epoch": 28.15536805353506, + "grad_norm": 0.12052540481090546, + "learning_rate": 2.9703192193605194e-06, + "loss": 0.0126, + "step": 96770 + }, + { + "epoch": 28.158277567646202, + "grad_norm": 0.08253193646669388, + "learning_rate": 2.967850414921879e-06, + "loss": 0.0121, + "step": 96780 + }, + { + "epoch": 28.161187081757348, + "grad_norm": 0.08992686867713928, + "learning_rate": 2.9653825242455923e-06, + "loss": 0.0104, + "step": 96790 + }, + { + "epoch": 28.16409659586849, + "grad_norm": 0.09221556037664413, + "learning_rate": 2.9629155475190866e-06, + "loss": 0.0108, + "step": 96800 + }, + { + "epoch": 28.167006109979635, + "grad_norm": 0.08842688053846359, + "learning_rate": 2.9604494849297038e-06, + "loss": 0.0136, + "step": 96810 + }, + { + "epoch": 28.169915624090777, + "grad_norm": 0.09163209795951843, + "learning_rate": 2.9579843366647267e-06, + "loss": 0.0102, + "step": 96820 + }, + { + "epoch": 28.17282513820192, + "grad_norm": 0.12358071655035019, + "learning_rate": 2.955520102911375e-06, + "loss": 0.0121, + "step": 96830 + }, + { + "epoch": 28.175734652313064, + "grad_norm": 0.14167384803295135, + "learning_rate": 2.953056783856783e-06, + "loss": 0.0135, + "step": 96840 + }, + { + "epoch": 28.178644166424206, + "grad_norm": 0.11418069154024124, + "learning_rate": 2.950594379688023e-06, + "loss": 0.013, + "step": 96850 + }, + { + "epoch": 28.18155368053535, + "grad_norm": 0.06615874916315079, + "learning_rate": 2.948132890592099e-06, + "loss": 0.0092, + "step": 96860 + }, + { + "epoch": 28.184463194646494, + "grad_norm": 0.157173290848732, + "learning_rate": 2.9456723167559515e-06, + "loss": 0.0119, + "step": 96870 + }, + { + "epoch": 28.18737270875764, + "grad_norm": 0.1299230009317398, + "learning_rate": 2.9432126583664354e-06, + "loss": 0.0105, + "step": 96880 + }, + { + "epoch": 28.19028222286878, + "grad_norm": 0.08996094018220901, + "learning_rate": 2.940753915610355e-06, + "loss": 0.0127, + "step": 96890 + }, + { + "epoch": 28.193191736979923, + "grad_norm": 0.22014160454273224, + "learning_rate": 2.9382960886744264e-06, + "loss": 0.0127, + "step": 96900 + }, + { + "epoch": 28.19610125109107, + "grad_norm": 0.0880308449268341, + "learning_rate": 2.935839177745308e-06, + "loss": 0.0101, + "step": 96910 + }, + { + "epoch": 28.19901076520221, + "grad_norm": 0.12600280344486237, + "learning_rate": 2.933383183009594e-06, + "loss": 0.0117, + "step": 96920 + }, + { + "epoch": 28.201920279313356, + "grad_norm": 0.08913987129926682, + "learning_rate": 2.930928104653789e-06, + "loss": 0.0106, + "step": 96930 + }, + { + "epoch": 28.204829793424498, + "grad_norm": 0.08044512569904327, + "learning_rate": 2.92847394286435e-06, + "loss": 0.0102, + "step": 96940 + }, + { + "epoch": 28.207739307535643, + "grad_norm": 0.3798866271972656, + "learning_rate": 2.9260206978276444e-06, + "loss": 0.0141, + "step": 96950 + }, + { + "epoch": 28.210648821646785, + "grad_norm": 0.10851485282182693, + "learning_rate": 2.923568369729992e-06, + "loss": 0.0114, + "step": 96960 + }, + { + "epoch": 28.213558335757927, + "grad_norm": 0.1453690528869629, + "learning_rate": 2.9211169587576186e-06, + "loss": 0.0105, + "step": 96970 + }, + { + "epoch": 28.216467849869073, + "grad_norm": 0.08750668168067932, + "learning_rate": 2.918666465096699e-06, + "loss": 0.0115, + "step": 96980 + }, + { + "epoch": 28.219377363980215, + "grad_norm": 0.10711812227964401, + "learning_rate": 2.916216888933335e-06, + "loss": 0.0127, + "step": 96990 + }, + { + "epoch": 28.22228687809136, + "grad_norm": 0.1936221718788147, + "learning_rate": 2.9137682304535524e-06, + "loss": 0.012, + "step": 97000 + }, + { + "epoch": 28.225196392202502, + "grad_norm": 0.2614877223968506, + "learning_rate": 2.911320489843303e-06, + "loss": 0.0118, + "step": 97010 + }, + { + "epoch": 28.228105906313644, + "grad_norm": 0.13195118308067322, + "learning_rate": 2.9088736672884923e-06, + "loss": 0.0112, + "step": 97020 + }, + { + "epoch": 28.23101542042479, + "grad_norm": 0.11671222746372223, + "learning_rate": 2.9064277629749315e-06, + "loss": 0.0143, + "step": 97030 + }, + { + "epoch": 28.23392493453593, + "grad_norm": 0.1294061243534088, + "learning_rate": 2.903982777088368e-06, + "loss": 0.0102, + "step": 97040 + }, + { + "epoch": 28.236834448647077, + "grad_norm": 0.1323583424091339, + "learning_rate": 2.901538709814486e-06, + "loss": 0.0098, + "step": 97050 + }, + { + "epoch": 28.23974396275822, + "grad_norm": 0.09344936907291412, + "learning_rate": 2.8990955613388994e-06, + "loss": 0.0121, + "step": 97060 + }, + { + "epoch": 28.242653476869364, + "grad_norm": 0.11015602201223373, + "learning_rate": 2.896653331847145e-06, + "loss": 0.0114, + "step": 97070 + }, + { + "epoch": 28.245562990980506, + "grad_norm": 0.12129401415586472, + "learning_rate": 2.8942120215246916e-06, + "loss": 0.0101, + "step": 97080 + }, + { + "epoch": 28.248472505091648, + "grad_norm": 0.12382204085588455, + "learning_rate": 2.891771630556945e-06, + "loss": 0.0124, + "step": 97090 + }, + { + "epoch": 28.251382019202794, + "grad_norm": 0.09607204794883728, + "learning_rate": 2.889332159129238e-06, + "loss": 0.0089, + "step": 97100 + }, + { + "epoch": 28.254291533313936, + "grad_norm": 0.11401520669460297, + "learning_rate": 2.8868936074268322e-06, + "loss": 0.0097, + "step": 97110 + }, + { + "epoch": 28.25720104742508, + "grad_norm": 0.06479570269584656, + "learning_rate": 2.8844559756349073e-06, + "loss": 0.0105, + "step": 97120 + }, + { + "epoch": 28.260110561536223, + "grad_norm": 0.09403529763221741, + "learning_rate": 2.882019263938605e-06, + "loss": 0.0103, + "step": 97130 + }, + { + "epoch": 28.26302007564737, + "grad_norm": 0.10944997519254684, + "learning_rate": 2.8795834725229663e-06, + "loss": 0.0101, + "step": 97140 + }, + { + "epoch": 28.26592958975851, + "grad_norm": 0.11742119491100311, + "learning_rate": 2.8771486015729733e-06, + "loss": 0.0117, + "step": 97150 + }, + { + "epoch": 28.268839103869652, + "grad_norm": 0.23130394518375397, + "learning_rate": 2.874714651273537e-06, + "loss": 0.0115, + "step": 97160 + }, + { + "epoch": 28.271748617980798, + "grad_norm": 0.17706042528152466, + "learning_rate": 2.8722816218095095e-06, + "loss": 0.0127, + "step": 97170 + }, + { + "epoch": 28.27465813209194, + "grad_norm": 0.08700937777757645, + "learning_rate": 2.869849513365653e-06, + "loss": 0.0115, + "step": 97180 + }, + { + "epoch": 28.277567646203085, + "grad_norm": 0.1070399284362793, + "learning_rate": 2.8674183261266764e-06, + "loss": 0.0103, + "step": 97190 + }, + { + "epoch": 28.280477160314227, + "grad_norm": 0.10712640732526779, + "learning_rate": 2.864988060277207e-06, + "loss": 0.0105, + "step": 97200 + }, + { + "epoch": 28.283386674425373, + "grad_norm": 0.08952908217906952, + "learning_rate": 2.8625587160018136e-06, + "loss": 0.011, + "step": 97210 + }, + { + "epoch": 28.286296188536515, + "grad_norm": 0.22029472887516022, + "learning_rate": 2.8601302934849827e-06, + "loss": 0.0106, + "step": 97220 + }, + { + "epoch": 28.289205702647656, + "grad_norm": 0.18591037392616272, + "learning_rate": 2.8577027929111387e-06, + "loss": 0.0127, + "step": 97230 + }, + { + "epoch": 28.292115216758802, + "grad_norm": 0.1436394900083542, + "learning_rate": 2.8552762144646394e-06, + "loss": 0.0108, + "step": 97240 + }, + { + "epoch": 28.295024730869944, + "grad_norm": 0.09955525398254395, + "learning_rate": 2.8528505583297603e-06, + "loss": 0.0111, + "step": 97250 + }, + { + "epoch": 28.29793424498109, + "grad_norm": 0.10290397703647614, + "learning_rate": 2.8504258246907193e-06, + "loss": 0.0133, + "step": 97260 + }, + { + "epoch": 28.30084375909223, + "grad_norm": 0.12144531309604645, + "learning_rate": 2.8480020137316544e-06, + "loss": 0.011, + "step": 97270 + }, + { + "epoch": 28.303753273203377, + "grad_norm": 0.16139477491378784, + "learning_rate": 2.845579125636639e-06, + "loss": 0.0131, + "step": 97280 + }, + { + "epoch": 28.30666278731452, + "grad_norm": 0.11221969872713089, + "learning_rate": 2.843157160589682e-06, + "loss": 0.0101, + "step": 97290 + }, + { + "epoch": 28.30957230142566, + "grad_norm": 0.12737680971622467, + "learning_rate": 2.8407361187747113e-06, + "loss": 0.0121, + "step": 97300 + }, + { + "epoch": 28.312481815536806, + "grad_norm": 0.084770567715168, + "learning_rate": 2.838316000375582e-06, + "loss": 0.0101, + "step": 97310 + }, + { + "epoch": 28.315391329647948, + "grad_norm": 0.12830707430839539, + "learning_rate": 2.835896805576093e-06, + "loss": 0.0132, + "step": 97320 + }, + { + "epoch": 28.318300843759094, + "grad_norm": 0.09640057384967804, + "learning_rate": 2.8334785345599704e-06, + "loss": 0.0129, + "step": 97330 + }, + { + "epoch": 28.321210357870235, + "grad_norm": 0.07978063821792603, + "learning_rate": 2.831061187510856e-06, + "loss": 0.0098, + "step": 97340 + }, + { + "epoch": 28.324119871981377, + "grad_norm": 0.14277233183383942, + "learning_rate": 2.8286447646123376e-06, + "loss": 0.0126, + "step": 97350 + }, + { + "epoch": 28.327029386092523, + "grad_norm": 0.13635756075382233, + "learning_rate": 2.8262292660479283e-06, + "loss": 0.0134, + "step": 97360 + }, + { + "epoch": 28.329938900203665, + "grad_norm": 0.1713048815727234, + "learning_rate": 2.8238146920010667e-06, + "loss": 0.0105, + "step": 97370 + }, + { + "epoch": 28.33284841431481, + "grad_norm": 0.1520623415708542, + "learning_rate": 2.8214010426551207e-06, + "loss": 0.0125, + "step": 97380 + }, + { + "epoch": 28.335757928425952, + "grad_norm": 0.13453158736228943, + "learning_rate": 2.8189883181933922e-06, + "loss": 0.0117, + "step": 97390 + }, + { + "epoch": 28.338667442537098, + "grad_norm": 0.08837773650884628, + "learning_rate": 2.8165765187991167e-06, + "loss": 0.0102, + "step": 97400 + }, + { + "epoch": 28.34157695664824, + "grad_norm": 0.06950005143880844, + "learning_rate": 2.814165644655452e-06, + "loss": 0.0102, + "step": 97410 + }, + { + "epoch": 28.34448647075938, + "grad_norm": 0.15676435828208923, + "learning_rate": 2.8117556959454787e-06, + "loss": 0.0102, + "step": 97420 + }, + { + "epoch": 28.347395984870527, + "grad_norm": 0.11227737367153168, + "learning_rate": 2.809346672852234e-06, + "loss": 0.0103, + "step": 97430 + }, + { + "epoch": 28.35030549898167, + "grad_norm": 0.10188087075948715, + "learning_rate": 2.8069385755586572e-06, + "loss": 0.0118, + "step": 97440 + }, + { + "epoch": 28.353215013092814, + "grad_norm": 0.11687441915273666, + "learning_rate": 2.804531404247625e-06, + "loss": 0.0116, + "step": 97450 + }, + { + "epoch": 28.356124527203956, + "grad_norm": 0.08592475950717926, + "learning_rate": 2.8021251591019486e-06, + "loss": 0.0097, + "step": 97460 + }, + { + "epoch": 28.359034041315102, + "grad_norm": 0.13374298810958862, + "learning_rate": 2.7997198403043724e-06, + "loss": 0.0103, + "step": 97470 + }, + { + "epoch": 28.361943555426244, + "grad_norm": 0.09212581813335419, + "learning_rate": 2.7973154480375545e-06, + "loss": 0.0123, + "step": 97480 + }, + { + "epoch": 28.364853069537386, + "grad_norm": 0.27165988087654114, + "learning_rate": 2.794911982484101e-06, + "loss": 0.0103, + "step": 97490 + }, + { + "epoch": 28.36776258364853, + "grad_norm": 0.178179532289505, + "learning_rate": 2.792509443826533e-06, + "loss": 0.0115, + "step": 97500 + }, + { + "epoch": 28.370672097759673, + "grad_norm": 0.5642987489700317, + "learning_rate": 2.790107832247314e-06, + "loss": 0.0099, + "step": 97510 + }, + { + "epoch": 28.37358161187082, + "grad_norm": 0.1536235511302948, + "learning_rate": 2.7877071479288248e-06, + "loss": 0.0109, + "step": 97520 + }, + { + "epoch": 28.37649112598196, + "grad_norm": 0.14793570339679718, + "learning_rate": 2.785307391053381e-06, + "loss": 0.0111, + "step": 97530 + }, + { + "epoch": 28.379400640093106, + "grad_norm": 0.09213881939649582, + "learning_rate": 2.7829085618032355e-06, + "loss": 0.0115, + "step": 97540 + }, + { + "epoch": 28.382310154204248, + "grad_norm": 0.08098451048135757, + "learning_rate": 2.780510660360556e-06, + "loss": 0.0109, + "step": 97550 + }, + { + "epoch": 28.38521966831539, + "grad_norm": 0.14281107485294342, + "learning_rate": 2.7781136869074535e-06, + "loss": 0.0113, + "step": 97560 + }, + { + "epoch": 28.388129182426535, + "grad_norm": 0.11939889937639236, + "learning_rate": 2.775717641625954e-06, + "loss": 0.0116, + "step": 97570 + }, + { + "epoch": 28.391038696537677, + "grad_norm": 0.08694462478160858, + "learning_rate": 2.7733225246980315e-06, + "loss": 0.0101, + "step": 97580 + }, + { + "epoch": 28.393948210648823, + "grad_norm": 0.08746728301048279, + "learning_rate": 2.770928336305571e-06, + "loss": 0.0119, + "step": 97590 + }, + { + "epoch": 28.396857724759965, + "grad_norm": 0.11682799458503723, + "learning_rate": 2.7685350766304e-06, + "loss": 0.0107, + "step": 97600 + }, + { + "epoch": 28.39976723887111, + "grad_norm": 0.13480263948440552, + "learning_rate": 2.766142745854267e-06, + "loss": 0.0109, + "step": 97610 + }, + { + "epoch": 28.402676752982252, + "grad_norm": 0.09539345651865005, + "learning_rate": 2.763751344158856e-06, + "loss": 0.0093, + "step": 97620 + }, + { + "epoch": 28.405586267093394, + "grad_norm": 0.09372766315937042, + "learning_rate": 2.7613608717257827e-06, + "loss": 0.0095, + "step": 97630 + }, + { + "epoch": 28.40849578120454, + "grad_norm": 0.06015473231673241, + "learning_rate": 2.7589713287365794e-06, + "loss": 0.0095, + "step": 97640 + }, + { + "epoch": 28.41140529531568, + "grad_norm": 0.07191997766494751, + "learning_rate": 2.756582715372719e-06, + "loss": 0.0129, + "step": 97650 + }, + { + "epoch": 28.414314809426827, + "grad_norm": 0.1006719097495079, + "learning_rate": 2.7541950318156056e-06, + "loss": 0.0111, + "step": 97660 + }, + { + "epoch": 28.41722432353797, + "grad_norm": 0.09715109318494797, + "learning_rate": 2.7518082782465635e-06, + "loss": 0.012, + "step": 97670 + }, + { + "epoch": 28.42013383764911, + "grad_norm": 0.12458264827728271, + "learning_rate": 2.74942245484685e-06, + "loss": 0.0096, + "step": 97680 + }, + { + "epoch": 28.423043351760256, + "grad_norm": 0.11561909317970276, + "learning_rate": 2.747037561797653e-06, + "loss": 0.0099, + "step": 97690 + }, + { + "epoch": 28.4259528658714, + "grad_norm": 0.095987968146801, + "learning_rate": 2.744653599280094e-06, + "loss": 0.0098, + "step": 97700 + }, + { + "epoch": 28.428862379982544, + "grad_norm": 0.1920228898525238, + "learning_rate": 2.7422705674752156e-06, + "loss": 0.0126, + "step": 97710 + }, + { + "epoch": 28.431771894093686, + "grad_norm": 0.09745332598686218, + "learning_rate": 2.7398884665639877e-06, + "loss": 0.0106, + "step": 97720 + }, + { + "epoch": 28.43468140820483, + "grad_norm": 0.10342087596654892, + "learning_rate": 2.7375072967273275e-06, + "loss": 0.0117, + "step": 97730 + }, + { + "epoch": 28.437590922315973, + "grad_norm": 0.10566645860671997, + "learning_rate": 2.735127058146063e-06, + "loss": 0.0097, + "step": 97740 + }, + { + "epoch": 28.440500436427115, + "grad_norm": 0.09446866810321808, + "learning_rate": 2.7327477510009536e-06, + "loss": 0.0111, + "step": 97750 + }, + { + "epoch": 28.44340995053826, + "grad_norm": 0.10791753232479095, + "learning_rate": 2.730369375472696e-06, + "loss": 0.0105, + "step": 97760 + }, + { + "epoch": 28.446319464649402, + "grad_norm": 0.08652960509061813, + "learning_rate": 2.727991931741916e-06, + "loss": 0.009, + "step": 97770 + }, + { + "epoch": 28.449228978760548, + "grad_norm": 0.12296925485134125, + "learning_rate": 2.725615419989159e-06, + "loss": 0.0126, + "step": 97780 + }, + { + "epoch": 28.45213849287169, + "grad_norm": 0.17461329698562622, + "learning_rate": 2.7232398403949043e-06, + "loss": 0.014, + "step": 97790 + }, + { + "epoch": 28.455048006982835, + "grad_norm": 0.10400432348251343, + "learning_rate": 2.720865193139564e-06, + "loss": 0.0091, + "step": 97800 + }, + { + "epoch": 28.457957521093977, + "grad_norm": 0.10892391949892044, + "learning_rate": 2.7184914784034804e-06, + "loss": 0.0094, + "step": 97810 + }, + { + "epoch": 28.46086703520512, + "grad_norm": 0.08354543894529343, + "learning_rate": 2.7161186963669145e-06, + "loss": 0.0121, + "step": 97820 + }, + { + "epoch": 28.463776549316265, + "grad_norm": 0.10773086547851562, + "learning_rate": 2.713746847210067e-06, + "loss": 0.0128, + "step": 97830 + }, + { + "epoch": 28.466686063427407, + "grad_norm": 0.14034081995487213, + "learning_rate": 2.7113759311130688e-06, + "loss": 0.0112, + "step": 97840 + }, + { + "epoch": 28.469595577538552, + "grad_norm": 0.07994251698255539, + "learning_rate": 2.709005948255966e-06, + "loss": 0.0085, + "step": 97850 + }, + { + "epoch": 28.472505091649694, + "grad_norm": 0.15023356676101685, + "learning_rate": 2.706636898818753e-06, + "loss": 0.0089, + "step": 97860 + }, + { + "epoch": 28.47541460576084, + "grad_norm": 0.10449380427598953, + "learning_rate": 2.7042687829813323e-06, + "loss": 0.0119, + "step": 97870 + }, + { + "epoch": 28.47832411987198, + "grad_norm": 0.130879208445549, + "learning_rate": 2.7019016009235588e-06, + "loss": 0.0106, + "step": 97880 + }, + { + "epoch": 28.481233633983123, + "grad_norm": 0.12427414953708649, + "learning_rate": 2.699535352825193e-06, + "loss": 0.0104, + "step": 97890 + }, + { + "epoch": 28.48414314809427, + "grad_norm": 0.11723575741052628, + "learning_rate": 2.6971700388659452e-06, + "loss": 0.0122, + "step": 97900 + }, + { + "epoch": 28.48705266220541, + "grad_norm": 0.10459046065807343, + "learning_rate": 2.694805659225438e-06, + "loss": 0.0112, + "step": 97910 + }, + { + "epoch": 28.489962176316556, + "grad_norm": 0.08155040442943573, + "learning_rate": 2.692442214083235e-06, + "loss": 0.0095, + "step": 97920 + }, + { + "epoch": 28.492871690427698, + "grad_norm": 0.08515691012144089, + "learning_rate": 2.690079703618825e-06, + "loss": 0.0113, + "step": 97930 + }, + { + "epoch": 28.495781204538844, + "grad_norm": 0.11430877447128296, + "learning_rate": 2.6877181280116203e-06, + "loss": 0.0092, + "step": 97940 + }, + { + "epoch": 28.498690718649986, + "grad_norm": 0.07788780331611633, + "learning_rate": 2.6853574874409736e-06, + "loss": 0.0087, + "step": 97950 + }, + { + "epoch": 28.501600232761128, + "grad_norm": 0.14842405915260315, + "learning_rate": 2.682997782086152e-06, + "loss": 0.012, + "step": 97960 + }, + { + "epoch": 28.504509746872273, + "grad_norm": 0.09185489267110825, + "learning_rate": 2.680639012126366e-06, + "loss": 0.0118, + "step": 97970 + }, + { + "epoch": 28.507419260983415, + "grad_norm": 0.22887377440929413, + "learning_rate": 2.678281177740744e-06, + "loss": 0.0121, + "step": 97980 + }, + { + "epoch": 28.51032877509456, + "grad_norm": 0.10567469894886017, + "learning_rate": 2.6759242791083518e-06, + "loss": 0.0103, + "step": 97990 + }, + { + "epoch": 28.513238289205702, + "grad_norm": 0.11717712134122849, + "learning_rate": 2.673568316408181e-06, + "loss": 0.0105, + "step": 98000 + }, + { + "epoch": 28.516147803316848, + "grad_norm": 0.10904695838689804, + "learning_rate": 2.671213289819149e-06, + "loss": 0.0119, + "step": 98010 + }, + { + "epoch": 28.51905731742799, + "grad_norm": 0.29598185420036316, + "learning_rate": 2.668859199520097e-06, + "loss": 0.0124, + "step": 98020 + }, + { + "epoch": 28.52196683153913, + "grad_norm": 0.09971896559000015, + "learning_rate": 2.666506045689818e-06, + "loss": 0.0128, + "step": 98030 + }, + { + "epoch": 28.524876345650277, + "grad_norm": 0.09877574443817139, + "learning_rate": 2.6641538285070107e-06, + "loss": 0.0114, + "step": 98040 + }, + { + "epoch": 28.52778585976142, + "grad_norm": 0.1170063465833664, + "learning_rate": 2.6618025481503077e-06, + "loss": 0.0109, + "step": 98050 + }, + { + "epoch": 28.530695373872565, + "grad_norm": 0.09255421161651611, + "learning_rate": 2.6594522047982756e-06, + "loss": 0.0103, + "step": 98060 + }, + { + "epoch": 28.533604887983707, + "grad_norm": 0.08552636951208115, + "learning_rate": 2.657102798629413e-06, + "loss": 0.0107, + "step": 98070 + }, + { + "epoch": 28.53651440209485, + "grad_norm": 0.12302565574645996, + "learning_rate": 2.654754329822135e-06, + "loss": 0.0092, + "step": 98080 + }, + { + "epoch": 28.539423916205994, + "grad_norm": 0.09225117415189743, + "learning_rate": 2.652406798554791e-06, + "loss": 0.0104, + "step": 98090 + }, + { + "epoch": 28.542333430317136, + "grad_norm": 0.10911740362644196, + "learning_rate": 2.650060205005664e-06, + "loss": 0.0096, + "step": 98100 + }, + { + "epoch": 28.54524294442828, + "grad_norm": 0.0957140177488327, + "learning_rate": 2.6477145493529652e-06, + "loss": 0.012, + "step": 98110 + }, + { + "epoch": 28.548152458539423, + "grad_norm": 0.09034126996994019, + "learning_rate": 2.6453698317748247e-06, + "loss": 0.0097, + "step": 98120 + }, + { + "epoch": 28.55106197265057, + "grad_norm": 0.10008101910352707, + "learning_rate": 2.6430260524493116e-06, + "loss": 0.0107, + "step": 98130 + }, + { + "epoch": 28.55397148676171, + "grad_norm": 0.07760139554738998, + "learning_rate": 2.6406832115544247e-06, + "loss": 0.0101, + "step": 98140 + }, + { + "epoch": 28.556881000872853, + "grad_norm": 0.08735030144453049, + "learning_rate": 2.6383413092680843e-06, + "loss": 0.0084, + "step": 98150 + }, + { + "epoch": 28.559790514983998, + "grad_norm": 0.1555217206478119, + "learning_rate": 2.6360003457681387e-06, + "loss": 0.0103, + "step": 98160 + }, + { + "epoch": 28.56270002909514, + "grad_norm": 0.083661749958992, + "learning_rate": 2.6336603212323693e-06, + "loss": 0.0104, + "step": 98170 + }, + { + "epoch": 28.565609543206286, + "grad_norm": 0.12828482687473297, + "learning_rate": 2.631321235838493e-06, + "loss": 0.0127, + "step": 98180 + }, + { + "epoch": 28.568519057317427, + "grad_norm": 0.09268027544021606, + "learning_rate": 2.6289830897641386e-06, + "loss": 0.0122, + "step": 98190 + }, + { + "epoch": 28.571428571428573, + "grad_norm": 0.15208229422569275, + "learning_rate": 2.6266458831868797e-06, + "loss": 0.0099, + "step": 98200 + }, + { + "epoch": 28.574338085539715, + "grad_norm": 0.11315120756626129, + "learning_rate": 2.624309616284207e-06, + "loss": 0.0111, + "step": 98210 + }, + { + "epoch": 28.577247599650857, + "grad_norm": 0.09908509254455566, + "learning_rate": 2.621974289233545e-06, + "loss": 0.0097, + "step": 98220 + }, + { + "epoch": 28.580157113762002, + "grad_norm": 0.1012888178229332, + "learning_rate": 2.619639902212251e-06, + "loss": 0.009, + "step": 98230 + }, + { + "epoch": 28.583066627873144, + "grad_norm": 0.09000766277313232, + "learning_rate": 2.6173064553976016e-06, + "loss": 0.0116, + "step": 98240 + }, + { + "epoch": 28.58597614198429, + "grad_norm": 0.10230209678411484, + "learning_rate": 2.6149739489668113e-06, + "loss": 0.0106, + "step": 98250 + }, + { + "epoch": 28.58888565609543, + "grad_norm": 0.0878186821937561, + "learning_rate": 2.6126423830970114e-06, + "loss": 0.0115, + "step": 98260 + }, + { + "epoch": 28.591795170206577, + "grad_norm": 0.08747760206460953, + "learning_rate": 2.6103117579652758e-06, + "loss": 0.0103, + "step": 98270 + }, + { + "epoch": 28.59470468431772, + "grad_norm": 0.10793249309062958, + "learning_rate": 2.607982073748596e-06, + "loss": 0.0107, + "step": 98280 + }, + { + "epoch": 28.59761419842886, + "grad_norm": 0.12229033559560776, + "learning_rate": 2.6056533306238955e-06, + "loss": 0.0104, + "step": 98290 + }, + { + "epoch": 28.600523712540006, + "grad_norm": 0.1567593216896057, + "learning_rate": 2.603325528768033e-06, + "loss": 0.0135, + "step": 98300 + }, + { + "epoch": 28.60343322665115, + "grad_norm": 0.12385322898626328, + "learning_rate": 2.6009986683577867e-06, + "loss": 0.0102, + "step": 98310 + }, + { + "epoch": 28.606342740762294, + "grad_norm": 0.10537239909172058, + "learning_rate": 2.598672749569861e-06, + "loss": 0.0093, + "step": 98320 + }, + { + "epoch": 28.609252254873436, + "grad_norm": 0.09883871674537659, + "learning_rate": 2.596347772580897e-06, + "loss": 0.0125, + "step": 98330 + }, + { + "epoch": 28.612161768984578, + "grad_norm": 0.11696377396583557, + "learning_rate": 2.5940237375674673e-06, + "loss": 0.0105, + "step": 98340 + }, + { + "epoch": 28.615071283095723, + "grad_norm": 0.11016026884317398, + "learning_rate": 2.5917006447060595e-06, + "loss": 0.013, + "step": 98350 + }, + { + "epoch": 28.617980797206865, + "grad_norm": 0.12948860228061676, + "learning_rate": 2.5893784941730995e-06, + "loss": 0.012, + "step": 98360 + }, + { + "epoch": 28.62089031131801, + "grad_norm": 0.05309295281767845, + "learning_rate": 2.5870572861449443e-06, + "loss": 0.0101, + "step": 98370 + }, + { + "epoch": 28.623799825429153, + "grad_norm": 0.11482275277376175, + "learning_rate": 2.5847370207978687e-06, + "loss": 0.009, + "step": 98380 + }, + { + "epoch": 28.626709339540298, + "grad_norm": 0.09720452129840851, + "learning_rate": 2.5824176983080804e-06, + "loss": 0.0105, + "step": 98390 + }, + { + "epoch": 28.62961885365144, + "grad_norm": 0.08250757306814194, + "learning_rate": 2.580099318851717e-06, + "loss": 0.0109, + "step": 98400 + }, + { + "epoch": 28.632528367762582, + "grad_norm": 0.113948754966259, + "learning_rate": 2.577781882604852e-06, + "loss": 0.0128, + "step": 98410 + }, + { + "epoch": 28.635437881873727, + "grad_norm": 0.1073017418384552, + "learning_rate": 2.575465389743472e-06, + "loss": 0.0119, + "step": 98420 + }, + { + "epoch": 28.63834739598487, + "grad_norm": 0.09532332420349121, + "learning_rate": 2.5731498404434945e-06, + "loss": 0.0106, + "step": 98430 + }, + { + "epoch": 28.641256910096015, + "grad_norm": 0.23106525838375092, + "learning_rate": 2.5708352348807835e-06, + "loss": 0.0106, + "step": 98440 + }, + { + "epoch": 28.644166424207157, + "grad_norm": 0.1134149432182312, + "learning_rate": 2.5685215732311103e-06, + "loss": 0.0095, + "step": 98450 + }, + { + "epoch": 28.647075938318302, + "grad_norm": 0.08352121710777283, + "learning_rate": 2.5662088556701807e-06, + "loss": 0.0091, + "step": 98460 + }, + { + "epoch": 28.649985452429444, + "grad_norm": 0.10607556998729706, + "learning_rate": 2.5638970823736312e-06, + "loss": 0.011, + "step": 98470 + }, + { + "epoch": 28.652894966540586, + "grad_norm": 0.08597410470247269, + "learning_rate": 2.5615862535170316e-06, + "loss": 0.0108, + "step": 98480 + }, + { + "epoch": 28.65580448065173, + "grad_norm": 0.10640964657068253, + "learning_rate": 2.559276369275866e-06, + "loss": 0.0117, + "step": 98490 + }, + { + "epoch": 28.658713994762874, + "grad_norm": 0.06963127851486206, + "learning_rate": 2.5569674298255625e-06, + "loss": 0.0112, + "step": 98500 + }, + { + "epoch": 28.66162350887402, + "grad_norm": 0.09400637447834015, + "learning_rate": 2.5546594353414625e-06, + "loss": 0.0108, + "step": 98510 + }, + { + "epoch": 28.66453302298516, + "grad_norm": 0.10921163856983185, + "learning_rate": 2.552352385998848e-06, + "loss": 0.0112, + "step": 98520 + }, + { + "epoch": 28.667442537096306, + "grad_norm": 0.12310309708118439, + "learning_rate": 2.5500462819729213e-06, + "loss": 0.0112, + "step": 98530 + }, + { + "epoch": 28.67035205120745, + "grad_norm": 0.1397581696510315, + "learning_rate": 2.5477411234388147e-06, + "loss": 0.012, + "step": 98540 + }, + { + "epoch": 28.67326156531859, + "grad_norm": 0.09340006858110428, + "learning_rate": 2.5454369105715965e-06, + "loss": 0.0112, + "step": 98550 + }, + { + "epoch": 28.676171079429736, + "grad_norm": 0.08621671050786972, + "learning_rate": 2.5431336435462494e-06, + "loss": 0.0107, + "step": 98560 + }, + { + "epoch": 28.679080593540878, + "grad_norm": 0.12417156994342804, + "learning_rate": 2.5408313225376955e-06, + "loss": 0.0111, + "step": 98570 + }, + { + "epoch": 28.681990107652023, + "grad_norm": 0.0848313719034195, + "learning_rate": 2.538529947720777e-06, + "loss": 0.0115, + "step": 98580 + }, + { + "epoch": 28.684899621763165, + "grad_norm": 0.1921319216489792, + "learning_rate": 2.53622951927027e-06, + "loss": 0.0118, + "step": 98590 + }, + { + "epoch": 28.68780913587431, + "grad_norm": 0.07869099825620651, + "learning_rate": 2.533930037360882e-06, + "loss": 0.0105, + "step": 98600 + }, + { + "epoch": 28.690718649985453, + "grad_norm": 0.09091523289680481, + "learning_rate": 2.5316315021672384e-06, + "loss": 0.0108, + "step": 98610 + }, + { + "epoch": 28.693628164096594, + "grad_norm": 0.09435641765594482, + "learning_rate": 2.5293339138638937e-06, + "loss": 0.0103, + "step": 98620 + }, + { + "epoch": 28.69653767820774, + "grad_norm": 0.10650788247585297, + "learning_rate": 2.5270372726253392e-06, + "loss": 0.0106, + "step": 98630 + }, + { + "epoch": 28.699447192318882, + "grad_norm": 0.08433486521244049, + "learning_rate": 2.5247415786259938e-06, + "loss": 0.0108, + "step": 98640 + }, + { + "epoch": 28.702356706430027, + "grad_norm": 0.11160383373498917, + "learning_rate": 2.5224468320401907e-06, + "loss": 0.0104, + "step": 98650 + }, + { + "epoch": 28.70526622054117, + "grad_norm": 0.07961607724428177, + "learning_rate": 2.5201530330422075e-06, + "loss": 0.0124, + "step": 98660 + }, + { + "epoch": 28.708175734652315, + "grad_norm": 0.12794265151023865, + "learning_rate": 2.517860181806245e-06, + "loss": 0.0112, + "step": 98670 + }, + { + "epoch": 28.711085248763457, + "grad_norm": 0.13509564101696014, + "learning_rate": 2.5155682785064266e-06, + "loss": 0.0118, + "step": 98680 + }, + { + "epoch": 28.7139947628746, + "grad_norm": 0.0991855338215828, + "learning_rate": 2.5132773233168006e-06, + "loss": 0.0109, + "step": 98690 + }, + { + "epoch": 28.716904276985744, + "grad_norm": 0.10036944597959518, + "learning_rate": 2.5109873164113623e-06, + "loss": 0.0126, + "step": 98700 + }, + { + "epoch": 28.719813791096886, + "grad_norm": 0.14113402366638184, + "learning_rate": 2.5086982579640177e-06, + "loss": 0.0109, + "step": 98710 + }, + { + "epoch": 28.72272330520803, + "grad_norm": 0.09047998487949371, + "learning_rate": 2.5064101481486062e-06, + "loss": 0.0096, + "step": 98720 + }, + { + "epoch": 28.725632819319173, + "grad_norm": 0.12691077589988708, + "learning_rate": 2.5041229871388854e-06, + "loss": 0.0135, + "step": 98730 + }, + { + "epoch": 28.72854233343032, + "grad_norm": 0.09234873205423355, + "learning_rate": 2.5018367751085647e-06, + "loss": 0.0098, + "step": 98740 + }, + { + "epoch": 28.73145184754146, + "grad_norm": 0.11564215272665024, + "learning_rate": 2.4995515122312616e-06, + "loss": 0.0103, + "step": 98750 + }, + { + "epoch": 28.734361361652603, + "grad_norm": 0.09374123066663742, + "learning_rate": 2.497267198680521e-06, + "loss": 0.0113, + "step": 98760 + }, + { + "epoch": 28.73727087576375, + "grad_norm": 0.1096271201968193, + "learning_rate": 2.494983834629826e-06, + "loss": 0.0097, + "step": 98770 + }, + { + "epoch": 28.74018038987489, + "grad_norm": 0.13074736297130585, + "learning_rate": 2.4927014202525856e-06, + "loss": 0.0098, + "step": 98780 + }, + { + "epoch": 28.743089903986036, + "grad_norm": 0.09975916147232056, + "learning_rate": 2.490419955722133e-06, + "loss": 0.0115, + "step": 98790 + }, + { + "epoch": 28.745999418097178, + "grad_norm": 0.16687904298305511, + "learning_rate": 2.48813944121172e-06, + "loss": 0.0108, + "step": 98800 + }, + { + "epoch": 28.74890893220832, + "grad_norm": 0.1188945472240448, + "learning_rate": 2.4858598768945523e-06, + "loss": 0.0101, + "step": 98810 + }, + { + "epoch": 28.751818446319465, + "grad_norm": 0.10018423944711685, + "learning_rate": 2.483581262943742e-06, + "loss": 0.011, + "step": 98820 + }, + { + "epoch": 28.754727960430607, + "grad_norm": 0.11326644569635391, + "learning_rate": 2.481303599532329e-06, + "loss": 0.0136, + "step": 98830 + }, + { + "epoch": 28.757637474541752, + "grad_norm": 0.1016308069229126, + "learning_rate": 2.4790268868332935e-06, + "loss": 0.0102, + "step": 98840 + }, + { + "epoch": 28.760546988652894, + "grad_norm": 0.09539709240198135, + "learning_rate": 2.476751125019536e-06, + "loss": 0.0102, + "step": 98850 + }, + { + "epoch": 28.76345650276404, + "grad_norm": 0.12865202128887177, + "learning_rate": 2.4744763142638817e-06, + "loss": 0.01, + "step": 98860 + }, + { + "epoch": 28.766366016875182, + "grad_norm": 0.11082407832145691, + "learning_rate": 2.4722024547390926e-06, + "loss": 0.0119, + "step": 98870 + }, + { + "epoch": 28.769275530986324, + "grad_norm": 0.10277251899242401, + "learning_rate": 2.4699295466178496e-06, + "loss": 0.0101, + "step": 98880 + }, + { + "epoch": 28.77218504509747, + "grad_norm": 0.11792582273483276, + "learning_rate": 2.4676575900727675e-06, + "loss": 0.0109, + "step": 98890 + }, + { + "epoch": 28.77509455920861, + "grad_norm": 0.08815564960241318, + "learning_rate": 2.4653865852763823e-06, + "loss": 0.0122, + "step": 98900 + }, + { + "epoch": 28.778004073319757, + "grad_norm": 0.09326225519180298, + "learning_rate": 2.463116532401166e-06, + "loss": 0.0116, + "step": 98910 + }, + { + "epoch": 28.7809135874309, + "grad_norm": 0.14395244419574738, + "learning_rate": 2.4608474316195144e-06, + "loss": 0.0106, + "step": 98920 + }, + { + "epoch": 28.783823101542044, + "grad_norm": 0.0867842584848404, + "learning_rate": 2.4585792831037463e-06, + "loss": 0.0108, + "step": 98930 + }, + { + "epoch": 28.786732615653186, + "grad_norm": 0.15091918408870697, + "learning_rate": 2.4563120870261195e-06, + "loss": 0.0127, + "step": 98940 + }, + { + "epoch": 28.789642129764328, + "grad_norm": 0.08363914489746094, + "learning_rate": 2.454045843558804e-06, + "loss": 0.0109, + "step": 98950 + }, + { + "epoch": 28.792551643875473, + "grad_norm": 0.11258859187364578, + "learning_rate": 2.4517805528739107e-06, + "loss": 0.0105, + "step": 98960 + }, + { + "epoch": 28.795461157986615, + "grad_norm": 0.14858020842075348, + "learning_rate": 2.449516215143478e-06, + "loss": 0.0102, + "step": 98970 + }, + { + "epoch": 28.79837067209776, + "grad_norm": 0.06824581325054169, + "learning_rate": 2.447252830539463e-06, + "loss": 0.0101, + "step": 98980 + }, + { + "epoch": 28.801280186208903, + "grad_norm": 0.10020966082811356, + "learning_rate": 2.44499039923375e-06, + "loss": 0.0088, + "step": 98990 + }, + { + "epoch": 28.804189700320048, + "grad_norm": 0.10217002034187317, + "learning_rate": 2.4427289213981602e-06, + "loss": 0.0117, + "step": 99000 + }, + { + "epoch": 28.80709921443119, + "grad_norm": 0.07805261760950089, + "learning_rate": 2.440468397204442e-06, + "loss": 0.012, + "step": 99010 + }, + { + "epoch": 28.810008728542332, + "grad_norm": 0.07987900823354721, + "learning_rate": 2.438208826824259e-06, + "loss": 0.0114, + "step": 99020 + }, + { + "epoch": 28.812918242653478, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.435950210429217e-06, + "loss": 0.0108, + "step": 99030 + }, + { + "epoch": 28.81582775676462, + "grad_norm": 0.11924725025892258, + "learning_rate": 2.4336925481908435e-06, + "loss": 0.0111, + "step": 99040 + }, + { + "epoch": 28.818737270875765, + "grad_norm": 0.09126931428909302, + "learning_rate": 2.43143584028059e-06, + "loss": 0.0093, + "step": 99050 + }, + { + "epoch": 28.821646784986907, + "grad_norm": 0.07232232391834259, + "learning_rate": 2.4291800868698354e-06, + "loss": 0.0115, + "step": 99060 + }, + { + "epoch": 28.82455629909805, + "grad_norm": 0.11351070553064346, + "learning_rate": 2.4269252881298936e-06, + "loss": 0.0106, + "step": 99070 + }, + { + "epoch": 28.827465813209194, + "grad_norm": 0.15333552658557892, + "learning_rate": 2.4246714442320057e-06, + "loss": 0.0107, + "step": 99080 + }, + { + "epoch": 28.830375327320336, + "grad_norm": 0.16162265837192535, + "learning_rate": 2.422418555347329e-06, + "loss": 0.0085, + "step": 99090 + }, + { + "epoch": 28.83328484143148, + "grad_norm": 0.10288070887327194, + "learning_rate": 2.4201666216469533e-06, + "loss": 0.011, + "step": 99100 + }, + { + "epoch": 28.836194355542624, + "grad_norm": 0.1820330023765564, + "learning_rate": 2.41791564330191e-06, + "loss": 0.0133, + "step": 99110 + }, + { + "epoch": 28.83910386965377, + "grad_norm": 0.09438538551330566, + "learning_rate": 2.4156656204831385e-06, + "loss": 0.0105, + "step": 99120 + }, + { + "epoch": 28.84201338376491, + "grad_norm": 0.0913604274392128, + "learning_rate": 2.4134165533615097e-06, + "loss": 0.0102, + "step": 99130 + }, + { + "epoch": 28.844922897876053, + "grad_norm": 0.1055363342165947, + "learning_rate": 2.4111684421078294e-06, + "loss": 0.0109, + "step": 99140 + }, + { + "epoch": 28.8478324119872, + "grad_norm": 0.11891163885593414, + "learning_rate": 2.4089212868928323e-06, + "loss": 0.0119, + "step": 99150 + }, + { + "epoch": 28.85074192609834, + "grad_norm": 0.0696154460310936, + "learning_rate": 2.4066750878871668e-06, + "loss": 0.0112, + "step": 99160 + }, + { + "epoch": 28.853651440209486, + "grad_norm": 0.0779939815402031, + "learning_rate": 2.4044298452614185e-06, + "loss": 0.0109, + "step": 99170 + }, + { + "epoch": 28.856560954320628, + "grad_norm": 0.08058398962020874, + "learning_rate": 2.4021855591860985e-06, + "loss": 0.0093, + "step": 99180 + }, + { + "epoch": 28.859470468431773, + "grad_norm": 0.1437763273715973, + "learning_rate": 2.3999422298316515e-06, + "loss": 0.0101, + "step": 99190 + }, + { + "epoch": 28.862379982542915, + "grad_norm": 0.1369076669216156, + "learning_rate": 2.3976998573684353e-06, + "loss": 0.0122, + "step": 99200 + }, + { + "epoch": 28.865289496654057, + "grad_norm": 0.12028560042381287, + "learning_rate": 2.3954584419667463e-06, + "loss": 0.0095, + "step": 99210 + }, + { + "epoch": 28.868199010765203, + "grad_norm": 0.10009239614009857, + "learning_rate": 2.3932179837968093e-06, + "loss": 0.0124, + "step": 99220 + }, + { + "epoch": 28.871108524876345, + "grad_norm": 0.16398055851459503, + "learning_rate": 2.3909784830287666e-06, + "loss": 0.013, + "step": 99230 + }, + { + "epoch": 28.87401803898749, + "grad_norm": 0.10879532247781754, + "learning_rate": 2.3887399398326976e-06, + "loss": 0.0126, + "step": 99240 + }, + { + "epoch": 28.876927553098632, + "grad_norm": 0.17434367537498474, + "learning_rate": 2.3865023543786007e-06, + "loss": 0.0119, + "step": 99250 + }, + { + "epoch": 28.879837067209778, + "grad_norm": 0.12708435952663422, + "learning_rate": 2.384265726836412e-06, + "loss": 0.0117, + "step": 99260 + }, + { + "epoch": 28.88274658132092, + "grad_norm": 0.09088437259197235, + "learning_rate": 2.382030057375981e-06, + "loss": 0.0096, + "step": 99270 + }, + { + "epoch": 28.88565609543206, + "grad_norm": 0.08715971559286118, + "learning_rate": 2.379795346167099e-06, + "loss": 0.0105, + "step": 99280 + }, + { + "epoch": 28.888565609543207, + "grad_norm": 0.08041825890541077, + "learning_rate": 2.3775615933794697e-06, + "loss": 0.01, + "step": 99290 + }, + { + "epoch": 28.89147512365435, + "grad_norm": 0.1375250518321991, + "learning_rate": 2.375328799182739e-06, + "loss": 0.0137, + "step": 99300 + }, + { + "epoch": 28.894384637765494, + "grad_norm": 0.12619927525520325, + "learning_rate": 2.3730969637464722e-06, + "loss": 0.0125, + "step": 99310 + }, + { + "epoch": 28.897294151876636, + "grad_norm": 0.11506721377372742, + "learning_rate": 2.370866087240158e-06, + "loss": 0.0105, + "step": 99320 + }, + { + "epoch": 28.90020366598778, + "grad_norm": 0.10955164581537247, + "learning_rate": 2.368636169833218e-06, + "loss": 0.0102, + "step": 99330 + }, + { + "epoch": 28.903113180098924, + "grad_norm": 0.08894036710262299, + "learning_rate": 2.3664072116950065e-06, + "loss": 0.01, + "step": 99340 + }, + { + "epoch": 28.906022694210066, + "grad_norm": 0.12434007227420807, + "learning_rate": 2.364179212994792e-06, + "loss": 0.0114, + "step": 99350 + }, + { + "epoch": 28.90893220832121, + "grad_norm": 0.10362933576107025, + "learning_rate": 2.361952173901773e-06, + "loss": 0.0114, + "step": 99360 + }, + { + "epoch": 28.911841722432353, + "grad_norm": 0.10673109441995621, + "learning_rate": 2.3597260945850842e-06, + "loss": 0.01, + "step": 99370 + }, + { + "epoch": 28.9147512365435, + "grad_norm": 0.12102537602186203, + "learning_rate": 2.3575009752137815e-06, + "loss": 0.0115, + "step": 99380 + }, + { + "epoch": 28.91766075065464, + "grad_norm": 0.35487326979637146, + "learning_rate": 2.3552768159568486e-06, + "loss": 0.0122, + "step": 99390 + }, + { + "epoch": 28.920570264765786, + "grad_norm": 0.09448620676994324, + "learning_rate": 2.3530536169831857e-06, + "loss": 0.0085, + "step": 99400 + }, + { + "epoch": 28.923479778876928, + "grad_norm": 0.12771224975585938, + "learning_rate": 2.3508313784616437e-06, + "loss": 0.0115, + "step": 99410 + }, + { + "epoch": 28.92638929298807, + "grad_norm": 0.095389224588871, + "learning_rate": 2.3486101005609828e-06, + "loss": 0.0126, + "step": 99420 + }, + { + "epoch": 28.929298807099215, + "grad_norm": 0.0789559856057167, + "learning_rate": 2.346389783449888e-06, + "loss": 0.0101, + "step": 99430 + }, + { + "epoch": 28.932208321210357, + "grad_norm": 0.09321089833974838, + "learning_rate": 2.344170427296985e-06, + "loss": 0.0103, + "step": 99440 + }, + { + "epoch": 28.935117835321503, + "grad_norm": 0.129488006234169, + "learning_rate": 2.3419520322708183e-06, + "loss": 0.0128, + "step": 99450 + }, + { + "epoch": 28.938027349432645, + "grad_norm": 0.07236841320991516, + "learning_rate": 2.33973459853986e-06, + "loss": 0.0121, + "step": 99460 + }, + { + "epoch": 28.940936863543786, + "grad_norm": 0.10917532444000244, + "learning_rate": 2.3375181262725036e-06, + "loss": 0.0122, + "step": 99470 + }, + { + "epoch": 28.943846377654932, + "grad_norm": 0.10473567247390747, + "learning_rate": 2.3353026156370816e-06, + "loss": 0.0093, + "step": 99480 + }, + { + "epoch": 28.946755891766074, + "grad_norm": 0.10416476428508759, + "learning_rate": 2.333088066801849e-06, + "loss": 0.0128, + "step": 99490 + }, + { + "epoch": 28.94966540587722, + "grad_norm": 0.1001383364200592, + "learning_rate": 2.33087447993498e-06, + "loss": 0.0118, + "step": 99500 + }, + { + "epoch": 28.95257491998836, + "grad_norm": 0.08039190620183945, + "learning_rate": 2.3286618552045856e-06, + "loss": 0.0092, + "step": 99510 + }, + { + "epoch": 28.955484434099507, + "grad_norm": 0.13723987340927124, + "learning_rate": 2.326450192778703e-06, + "loss": 0.0122, + "step": 99520 + }, + { + "epoch": 28.95839394821065, + "grad_norm": 0.11497245728969574, + "learning_rate": 2.3242394928252903e-06, + "loss": 0.0122, + "step": 99530 + }, + { + "epoch": 28.96130346232179, + "grad_norm": 0.15377052128314972, + "learning_rate": 2.3220297555122317e-06, + "loss": 0.0124, + "step": 99540 + }, + { + "epoch": 28.964212976432936, + "grad_norm": 0.14190872013568878, + "learning_rate": 2.3198209810073463e-06, + "loss": 0.0114, + "step": 99550 + }, + { + "epoch": 28.967122490544078, + "grad_norm": 0.101814866065979, + "learning_rate": 2.3176131694783782e-06, + "loss": 0.0104, + "step": 99560 + }, + { + "epoch": 28.970032004655224, + "grad_norm": 0.10409971326589584, + "learning_rate": 2.31540632109299e-06, + "loss": 0.0109, + "step": 99570 + }, + { + "epoch": 28.972941518766365, + "grad_norm": 0.142991840839386, + "learning_rate": 2.313200436018786e-06, + "loss": 0.0109, + "step": 99580 + }, + { + "epoch": 28.97585103287751, + "grad_norm": 0.12948636710643768, + "learning_rate": 2.3109955144232768e-06, + "loss": 0.0093, + "step": 99590 + }, + { + "epoch": 28.978760546988653, + "grad_norm": 0.09271764010190964, + "learning_rate": 2.308791556473921e-06, + "loss": 0.0104, + "step": 99600 + }, + { + "epoch": 28.981670061099795, + "grad_norm": 0.10465230792760849, + "learning_rate": 2.306588562338094e-06, + "loss": 0.0137, + "step": 99610 + }, + { + "epoch": 28.98457957521094, + "grad_norm": 0.14591771364212036, + "learning_rate": 2.304386532183092e-06, + "loss": 0.0142, + "step": 99620 + }, + { + "epoch": 28.987489089322082, + "grad_norm": 0.09760881960391998, + "learning_rate": 2.3021854661761534e-06, + "loss": 0.0125, + "step": 99630 + }, + { + "epoch": 28.990398603433228, + "grad_norm": 0.11471458524465561, + "learning_rate": 2.299985364484427e-06, + "loss": 0.0095, + "step": 99640 + }, + { + "epoch": 28.99330811754437, + "grad_norm": 0.09893053770065308, + "learning_rate": 2.297786227275003e-06, + "loss": 0.0102, + "step": 99650 + }, + { + "epoch": 28.996217631655515, + "grad_norm": 0.12702278792858124, + "learning_rate": 2.2955880547148833e-06, + "loss": 0.0111, + "step": 99660 + }, + { + "epoch": 28.999127145766657, + "grad_norm": 0.14084993302822113, + "learning_rate": 2.29339084697101e-06, + "loss": 0.0098, + "step": 99670 + }, + { + "epoch": 29.0020366598778, + "grad_norm": 0.09953160583972931, + "learning_rate": 2.291194604210249e-06, + "loss": 0.0111, + "step": 99680 + }, + { + "epoch": 29.004946173988944, + "grad_norm": 0.10567528754472733, + "learning_rate": 2.2889993265993875e-06, + "loss": 0.0088, + "step": 99690 + }, + { + "epoch": 29.007855688100086, + "grad_norm": 0.11675871163606644, + "learning_rate": 2.2868050143051343e-06, + "loss": 0.0114, + "step": 99700 + }, + { + "epoch": 29.010765202211232, + "grad_norm": 0.09852224588394165, + "learning_rate": 2.2846116674941465e-06, + "loss": 0.0115, + "step": 99710 + }, + { + "epoch": 29.013674716322374, + "grad_norm": 0.09203079342842102, + "learning_rate": 2.28241928633299e-06, + "loss": 0.0105, + "step": 99720 + }, + { + "epoch": 29.01658423043352, + "grad_norm": 0.12082663923501968, + "learning_rate": 2.2802278709881557e-06, + "loss": 0.0122, + "step": 99730 + }, + { + "epoch": 29.01949374454466, + "grad_norm": 0.10222533345222473, + "learning_rate": 2.27803742162607e-06, + "loss": 0.0091, + "step": 99740 + }, + { + "epoch": 29.022403258655803, + "grad_norm": 0.11994101852178574, + "learning_rate": 2.2758479384130903e-06, + "loss": 0.0106, + "step": 99750 + }, + { + "epoch": 29.02531277276695, + "grad_norm": 0.10856207460165024, + "learning_rate": 2.2736594215154865e-06, + "loss": 0.0092, + "step": 99760 + }, + { + "epoch": 29.02822228687809, + "grad_norm": 0.08752094954252243, + "learning_rate": 2.2714718710994594e-06, + "loss": 0.0115, + "step": 99770 + }, + { + "epoch": 29.031131800989236, + "grad_norm": 0.1148502454161644, + "learning_rate": 2.269285287331144e-06, + "loss": 0.0104, + "step": 99780 + }, + { + "epoch": 29.034041315100378, + "grad_norm": 0.11640161275863647, + "learning_rate": 2.2670996703765973e-06, + "loss": 0.0103, + "step": 99790 + }, + { + "epoch": 29.03695082921152, + "grad_norm": 0.18910947442054749, + "learning_rate": 2.264915020401799e-06, + "loss": 0.0111, + "step": 99800 + }, + { + "epoch": 29.039860343322665, + "grad_norm": 0.143405944108963, + "learning_rate": 2.2627313375726587e-06, + "loss": 0.0121, + "step": 99810 + }, + { + "epoch": 29.042769857433807, + "grad_norm": 0.10507558286190033, + "learning_rate": 2.2605486220550196e-06, + "loss": 0.0098, + "step": 99820 + }, + { + "epoch": 29.045679371544953, + "grad_norm": 0.249084934592247, + "learning_rate": 2.258366874014639e-06, + "loss": 0.0103, + "step": 99830 + }, + { + "epoch": 29.048588885656095, + "grad_norm": 0.09517933428287506, + "learning_rate": 2.256186093617204e-06, + "loss": 0.0128, + "step": 99840 + }, + { + "epoch": 29.05149839976724, + "grad_norm": 0.08562472462654114, + "learning_rate": 2.2540062810283326e-06, + "loss": 0.0106, + "step": 99850 + }, + { + "epoch": 29.054407913878382, + "grad_norm": 0.11253012716770172, + "learning_rate": 2.2518274364135716e-06, + "loss": 0.0115, + "step": 99860 + }, + { + "epoch": 29.057317427989524, + "grad_norm": 0.12640061974525452, + "learning_rate": 2.249649559938383e-06, + "loss": 0.01, + "step": 99870 + }, + { + "epoch": 29.06022694210067, + "grad_norm": 0.17042694985866547, + "learning_rate": 2.2474726517681683e-06, + "loss": 0.0095, + "step": 99880 + }, + { + "epoch": 29.06313645621181, + "grad_norm": 0.1118495836853981, + "learning_rate": 2.2452967120682444e-06, + "loss": 0.0096, + "step": 99890 + }, + { + "epoch": 29.066045970322957, + "grad_norm": 0.0998217910528183, + "learning_rate": 2.243121741003859e-06, + "loss": 0.0117, + "step": 99900 + }, + { + "epoch": 29.0689554844341, + "grad_norm": 0.08165398985147476, + "learning_rate": 2.240947738740195e-06, + "loss": 0.0096, + "step": 99910 + }, + { + "epoch": 29.071864998545244, + "grad_norm": 0.11531069874763489, + "learning_rate": 2.2387747054423436e-06, + "loss": 0.0116, + "step": 99920 + }, + { + "epoch": 29.074774512656386, + "grad_norm": 0.11632228642702103, + "learning_rate": 2.236602641275339e-06, + "loss": 0.0105, + "step": 99930 + }, + { + "epoch": 29.07768402676753, + "grad_norm": 0.13619601726531982, + "learning_rate": 2.2344315464041305e-06, + "loss": 0.0103, + "step": 99940 + }, + { + "epoch": 29.080593540878674, + "grad_norm": 0.13361766934394836, + "learning_rate": 2.232261420993603e-06, + "loss": 0.0118, + "step": 99950 + }, + { + "epoch": 29.083503054989816, + "grad_norm": 0.11400602012872696, + "learning_rate": 2.2300922652085592e-06, + "loss": 0.0106, + "step": 99960 + }, + { + "epoch": 29.08641256910096, + "grad_norm": 0.09594951570034027, + "learning_rate": 2.227924079213732e-06, + "loss": 0.0116, + "step": 99970 + }, + { + "epoch": 29.089322083212103, + "grad_norm": 0.12091630697250366, + "learning_rate": 2.2257568631737873e-06, + "loss": 0.0112, + "step": 99980 + }, + { + "epoch": 29.09223159732325, + "grad_norm": 0.11330015957355499, + "learning_rate": 2.223590617253306e-06, + "loss": 0.0118, + "step": 99990 + }, + { + "epoch": 29.09514111143439, + "grad_norm": 0.09932112693786621, + "learning_rate": 2.221425341616796e-06, + "loss": 0.0104, + "step": 100000 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 20000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 64, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-120000/config.json b/checkpoint-120000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d7b77e4255263f3803a4b6f94ca6831e459b70 --- /dev/null +++ b/checkpoint-120000/config.json @@ -0,0 +1,64 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/checkpoint-120000/experiment_cfg/metadata.json b/checkpoint-120000/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..300eb070963337016e16465f50ba8351ea2eb25c --- /dev/null +++ b/checkpoint-120000/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.3139495849609375, + 0.4876587688922882, + 0.7196521759033203 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6996827125549316 + ], + "mean": [ + 2.749288365724937, + -2.0455216239909983, + 0.700532551020533 + ], + "std": [ + 1.6786295897046262, + 1.312897969434244, + 0.00133751758906258 + ], + "q01": [ + -1.4241907881148037, + -5.1716547935950885, + 0.6999670898768614 + ], + "q99": [ + 6.198111545831555, + -0.5873531334104489, + 0.7038833014242587 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.23084999587450247, + 0.6238431088581847 + ], + "std": [ + 0.0, + 0.0, + 0.6556404371644047, + 0.3572252105732042 + ], + "q01": [ + 0.0, + 0.0, + -0.9999999999999999, + 1.7482354593273349e-06 + ], + "q99": [ + 0.0, + 0.0, + 0.9999999999999999, + 0.9999999999999999 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8003877401351929, + 0.9829942584037781 + ], + "min": [ + -0.37471598386764526, + -0.8472502827644348, + -0.25070279836654663 + ], + "mean": [ + 0.28667332601863704, + -0.038315473170876704, + 0.45974658906777444 + ], + "std": [ + 0.17067058828305154, + 0.23569952037784195, + 0.21950702354181487 + ], + "q01": [ + 0.015172948963784228, + -0.44450502063220615, + 0.23314444103329338 + ], + "q99": [ + 0.5609069689572412, + 0.38454179128009547, + 0.7028102736473852 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9984455108642578, + 0.9434727430343628, + 0.9062229990959167 + ], + "min": [ + -0.9999930262565613, + -0.9988337755203247, + -0.9618149995803833, + 2.1872274658107926e-07 + ], + "mean": [ + -0.25672297401336, + 0.024425335806687216, + -0.08740977164483847, + 0.16608739644018397 + ], + "std": [ + 0.7932585208392383, + 0.3102260195580416, + 0.3772472576315148, + 0.17451959300158773 + ], + "q01": [ + -0.99379006987882, + -0.5478102050038828, + -0.6101777952971636, + 0.002274085014134748 + ], + "q99": [ + 0.8804856038896288, + 0.5910611092873037, + 0.5216058042993562, + 0.5231965974012255 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011436971835792065, + -0.05664053186774254 + ], + "mean": [ + 0.031651551516052555, + -0.03161481387920421 + ], + "std": [ + 0.013119769222217526, + 0.01306078225192402 + ], + "q01": [ + 0.0060999246446777336, + -0.04062601653964198 + ], + "q99": [ + 0.04054891621517283, + -0.0059163567113456345 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.008144896753718373, + -0.00018893135146078263, + -0.0008739062727221845, + 0.0 + ], + "std": [ + 0.11030045424364411, + 0.10148082570313594, + 0.08975698373467506, + 0.0 + ], + "q01": [ + -0.07287373067581518, + -0.09446994199569948, + -0.07702818259249572, + 0.0 + ], + "q99": [ + 0.04485485414407898, + 0.09626941605965177, + 0.07610665906122742, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9221974313425939 + ], + "std": [ + 0.38670194342612674 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.384693946883099 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.01976129923017491, + -0.020870631344490034, + -0.05993476517349181 + ], + "std": [ + 0.4347024990804053, + 0.4221662976569997, + 0.3845506843044066 + ], + "q01": [ + -0.8835792939767807, + -0.9280919251713778, + -0.783361071470956 + ], + "q99": [ + 0.7622084438449307, + 0.8625125252609077, + 0.7470974753250706 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.008089673068544335, + -0.026498254627927348, + 0.0029083551743059005 + ], + "std": [ + 0.11216734503733773, + 0.13206002613798629, + 0.12615870412692864 + ], + "q01": [ + -0.2814627972846672, + -0.4342386516115439, + -0.34489755628340574 + ], + "q99": [ + 0.31879353266939386, + 0.28411797683220563, + 0.3597718671669894 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.3824228874769045 + ], + "std": [ + 0.9240015096469786 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.5597272305813592 + ] + } + } + }, + "modalities": { + "video": { + "robot0_eye_in_hand": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_left": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_right": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": true, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + }, + "end_effector_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": true, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/checkpoint-120000/model-00001-of-00002.safetensors b/checkpoint-120000/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d7d8c2b55704776186a9207b3175b6c657758e2 --- /dev/null +++ b/checkpoint-120000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9fc9945eca28b94cc6398098b3d2092b0e9731d322853d46197bf3dc7a37a961 +size 4999367032 diff --git a/checkpoint-120000/model-00002-of-00002.safetensors b/checkpoint-120000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..42080be5aa6a552d612cad02b976aaea03277ce1 --- /dev/null +++ b/checkpoint-120000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be562fdffb28a9a991de2c352b351bae7ac8e29e18c5db5b96a5fa82c26b8e39 +size 2586705312 diff --git a/checkpoint-120000/model.safetensors.index.json b/checkpoint-120000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/checkpoint-120000/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-120000/optimizer.pt b/checkpoint-120000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..a702fe4b5677b68461a02f443908b0092aa062f5 --- /dev/null +++ b/checkpoint-120000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:308126cd456c9643bd7ed6bcd83fea62fa69173101e6d24dc85bd16713700d7a +size 8550719959 diff --git a/checkpoint-120000/rng_state_0.pth b/checkpoint-120000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..d023dc8b0f5545f61efbff503e315887eb40cbbc --- /dev/null +++ b/checkpoint-120000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:aaee3f87d86ca7c8f6c5905597cfb2c789c73107a5e25bd09b8a8f23b09fdd58 +size 14917 diff --git a/checkpoint-120000/rng_state_1.pth b/checkpoint-120000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..61636dc4acc5d1dd521477f267a24d266d76f69c --- /dev/null +++ b/checkpoint-120000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:eb3344198659f488c4af3a8a0d404c95eca67d581a6aa90b492931af2279900c +size 14917 diff --git a/checkpoint-120000/scheduler.pt b/checkpoint-120000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..575e03349720727e462b57642c71591e2b061dac --- /dev/null +++ b/checkpoint-120000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:48f412abc82c519159c43890e65e38c5467c83b8c0467a1cc8ef436b8360d71f +size 1465 diff --git a/checkpoint-120000/trainer_state.json b/checkpoint-120000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..3b489a9688acfbb25e8da771c840f4704222c8db --- /dev/null +++ b/checkpoint-120000/trainer_state.json @@ -0,0 +1,84034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 34.91416933372127, + "eval_steps": 500, + "global_step": 120000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002909514111143439, + "grad_norm": 5.5908684730529785, + "learning_rate": 4.5e-08, + "loss": 0.5668, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.084368705749512, + "learning_rate": 9.5e-08, + "loss": 0.5781, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 5.755157947540283, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5759, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.174816608428955, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5202, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 6.199656963348389, + "learning_rate": 2.45e-07, + "loss": 0.5455, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.251872539520264, + "learning_rate": 2.95e-07, + "loss": 0.5507, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 5.606583595275879, + "learning_rate": 3.45e-07, + "loss": 0.5264, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.967868328094482, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.5166, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.92909574508667, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4578, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.306433200836182, + "learning_rate": 4.95e-07, + "loss": 0.4091, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.75846529006958, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4072, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.2030856609344482, + "learning_rate": 5.95e-07, + "loss": 0.2992, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.5756216049194336, + "learning_rate": 6.45e-07, + "loss": 0.2703, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.3420923948287964, + "learning_rate": 6.95e-07, + "loss": 0.2492, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 0.9191234111785889, + "learning_rate": 7.45e-07, + "loss": 0.2302, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.662022054195404, + "learning_rate": 7.95e-07, + "loss": 0.1993, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.704409658908844, + "learning_rate": 8.45e-07, + "loss": 0.1814, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.6104884147644043, + "learning_rate": 8.95e-07, + "loss": 0.1694, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.4955693781375885, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1674, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.707667887210846, + "learning_rate": 9.95e-07, + "loss": 0.1507, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.6977017521858215, + "learning_rate": 1.045e-06, + "loss": 0.1547, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.49085867404937744, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1536, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.6363996267318726, + "learning_rate": 1.145e-06, + "loss": 0.1565, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.4633055627346039, + "learning_rate": 1.195e-06, + "loss": 0.1641, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5155240297317505, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1513, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5815547108650208, + "learning_rate": 1.295e-06, + "loss": 0.1431, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.5865835547447205, + "learning_rate": 1.345e-06, + "loss": 0.1419, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.5872474312782288, + "learning_rate": 1.395e-06, + "loss": 0.1326, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.4897519052028656, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1326, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.46316635608673096, + "learning_rate": 1.495e-06, + "loss": 0.1442, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.5359047651290894, + "learning_rate": 1.545e-06, + "loss": 0.1394, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.5805708169937134, + "learning_rate": 1.595e-06, + "loss": 0.1231, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.5488415360450745, + "learning_rate": 1.645e-06, + "loss": 0.127, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.41755908727645874, + "learning_rate": 1.695e-06, + "loss": 0.1375, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.6258692145347595, + "learning_rate": 1.745e-06, + "loss": 0.1202, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.5017617344856262, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1317, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.710078239440918, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.46977895498275757, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1215, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.4802077114582062, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.116, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.6448793411254883, + "learning_rate": 1.995e-06, + "loss": 0.1172, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4394509494304657, + "learning_rate": 2.045e-06, + "loss": 0.1145, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.6547806859016418, + "learning_rate": 2.095e-06, + "loss": 0.1174, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.6954914331436157, + "learning_rate": 2.145e-06, + "loss": 0.1245, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.670100212097168, + "learning_rate": 2.195e-06, + "loss": 0.1133, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.5230806469917297, + "learning_rate": 2.245e-06, + "loss": 0.11, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.46809589862823486, + "learning_rate": 2.295e-06, + "loss": 0.1069, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.6178835034370422, + "learning_rate": 2.345e-06, + "loss": 0.1099, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.41567298769950867, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1072, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4968737065792084, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1093, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.46893924474716187, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1046, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.5041665434837341, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1014, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5833673477172852, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1095, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.5565881133079529, + "learning_rate": 2.645e-06, + "loss": 0.1057, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6213339567184448, + "learning_rate": 2.695e-06, + "loss": 0.1087, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.5638794898986816, + "learning_rate": 2.745e-06, + "loss": 0.099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.5594047904014587, + "learning_rate": 2.795e-06, + "loss": 0.1065, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5521039366722107, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1072, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.5013270378112793, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1045, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.44355231523513794, + "learning_rate": 2.945e-06, + "loss": 0.0985, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.5467342734336853, + "learning_rate": 2.995e-06, + "loss": 0.101, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5065611600875854, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0939, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.5587425827980042, + "learning_rate": 3.095e-06, + "loss": 0.0969, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5561433434486389, + "learning_rate": 3.145e-06, + "loss": 0.1009, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.5576680302619934, + "learning_rate": 3.195e-06, + "loss": 0.1006, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.7426344752311707, + "learning_rate": 3.245e-06, + "loss": 0.0959, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.64913409948349, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.0962, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.5271337628364563, + "learning_rate": 3.345e-06, + "loss": 0.0927, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.6992810368537903, + "learning_rate": 3.395e-06, + "loss": 0.0924, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.559481680393219, + "learning_rate": 3.445e-06, + "loss": 0.0886, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7303935885429382, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0918, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.45654281973838806, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0934, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.41098082065582275, + "learning_rate": 3.595e-06, + "loss": 0.0911, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.43506675958633423, + "learning_rate": 3.645e-06, + "loss": 0.0887, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5131197571754456, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0983, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.4571124017238617, + "learning_rate": 3.745e-06, + "loss": 0.0928, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.6749465465545654, + "learning_rate": 3.795e-06, + "loss": 0.0926, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.46945369243621826, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1016, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.6481541395187378, + "learning_rate": 3.895e-06, + "loss": 0.0917, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.6769195795059204, + "learning_rate": 3.945e-06, + "loss": 0.0924, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5225585103034973, + "learning_rate": 3.995e-06, + "loss": 0.0909, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.642867922782898, + "learning_rate": 4.045e-06, + "loss": 0.0829, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.4617709219455719, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0924, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.5350925326347351, + "learning_rate": 4.145e-06, + "loss": 0.0821, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.4871961176395416, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0896, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.5545480847358704, + "learning_rate": 4.245e-06, + "loss": 0.0891, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.6057930588722229, + "learning_rate": 4.295e-06, + "loss": 0.0876, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.515282154083252, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0873, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.5402254462242126, + "learning_rate": 4.395e-06, + "loss": 0.0908, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4669134318828583, + "learning_rate": 4.445e-06, + "loss": 0.0837, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.46819189190864563, + "learning_rate": 4.495e-06, + "loss": 0.0888, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35766035318374634, + "learning_rate": 4.545e-06, + "loss": 0.0829, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.6769008040428162, + "learning_rate": 4.595e-06, + "loss": 0.0817, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5884619355201721, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0892, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5613101720809937, + "learning_rate": 4.695e-06, + "loss": 0.0852, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.47087088227272034, + "learning_rate": 4.745e-06, + "loss": 0.0818, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.4887458384037018, + "learning_rate": 4.795e-06, + "loss": 0.0858, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.44256526231765747, + "learning_rate": 4.845e-06, + "loss": 0.0833, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5375688076019287, + "learning_rate": 4.895e-06, + "loss": 0.075, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.6297019124031067, + "learning_rate": 4.945e-06, + "loss": 0.0932, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.6576125621795654, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0829, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.724773108959198, + "learning_rate": 5.045e-06, + "loss": 0.0813, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.54575115442276, + "learning_rate": 5.095e-06, + "loss": 0.0873, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.5186223387718201, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0736, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.46651381254196167, + "learning_rate": 5.195e-06, + "loss": 0.0791, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.5323564410209656, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0753, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.42579638957977295, + "learning_rate": 5.295e-06, + "loss": 0.0755, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.5063076019287109, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0767, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.4558819532394409, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0807, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.6920697093009949, + "learning_rate": 5.445e-06, + "loss": 0.079, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.4706108868122101, + "learning_rate": 5.495e-06, + "loss": 0.079, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.46705493330955505, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0745, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4643767774105072, + "learning_rate": 5.595e-06, + "loss": 0.0778, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.4800759553909302, + "learning_rate": 5.645e-06, + "loss": 0.0792, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.37705323100090027, + "learning_rate": 5.695e-06, + "loss": 0.0815, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.6651142239570618, + "learning_rate": 5.745e-06, + "loss": 0.0732, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.5275036692619324, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0758, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.45047008991241455, + "learning_rate": 5.845e-06, + "loss": 0.0808, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.6931971311569214, + "learning_rate": 5.895e-06, + "loss": 0.0808, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.7053916454315186, + "learning_rate": 5.945e-06, + "loss": 0.0747, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.6798260807991028, + "learning_rate": 5.995e-06, + "loss": 0.0829, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.5724003314971924, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0749, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.5836488008499146, + "learning_rate": 6.095e-06, + "loss": 0.0759, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.40817850828170776, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0778, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.41070470213890076, + "learning_rate": 6.195e-06, + "loss": 0.0765, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.43642961978912354, + "learning_rate": 6.245e-06, + "loss": 0.0711, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.49526578187942505, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0713, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.7773705720901489, + "learning_rate": 6.345e-06, + "loss": 0.0744, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.5485936999320984, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0732, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.5593521595001221, + "learning_rate": 6.445e-06, + "loss": 0.0747, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.3667270839214325, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0723, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.5555917620658875, + "learning_rate": 6.545e-06, + "loss": 0.0726, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.38107141852378845, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0647, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.46452224254608154, + "learning_rate": 6.645e-06, + "loss": 0.0787, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6018183827400208, + "learning_rate": 6.695e-06, + "loss": 0.0718, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.44997578859329224, + "learning_rate": 6.745e-06, + "loss": 0.0758, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45717573165893555, + "learning_rate": 6.795e-06, + "loss": 0.0738, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.5185182690620422, + "learning_rate": 6.845e-06, + "loss": 0.0768, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.5135819315910339, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.43529245257377625, + "learning_rate": 6.945e-06, + "loss": 0.0691, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.49389752745628357, + "learning_rate": 6.995e-06, + "loss": 0.0678, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.4113156199455261, + "learning_rate": 7.045e-06, + "loss": 0.0651, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.4072611927986145, + "learning_rate": 7.095e-06, + "loss": 0.0691, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.4524228870868683, + "learning_rate": 7.145e-06, + "loss": 0.0688, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.6225605607032776, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0676, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4841763377189636, + "learning_rate": 7.245e-06, + "loss": 0.0678, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.46297937631607056, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0759, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.5585324168205261, + "learning_rate": 7.345e-06, + "loss": 0.0671, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.57328200340271, + "learning_rate": 7.395e-06, + "loss": 0.0717, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.5072467923164368, + "learning_rate": 7.445000000000001e-06, + "loss": 0.071, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.4630240797996521, + "learning_rate": 7.495e-06, + "loss": 0.0674, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.35908272862434387, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0704, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.5117342472076416, + "learning_rate": 7.595e-06, + "loss": 0.0675, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.5214630961418152, + "learning_rate": 7.645e-06, + "loss": 0.0702, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.39036089181900024, + "learning_rate": 7.695e-06, + "loss": 0.0709, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.558722734451294, + "learning_rate": 7.745e-06, + "loss": 0.0705, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.3612176179885864, + "learning_rate": 7.795e-06, + "loss": 0.0715, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.523444652557373, + "learning_rate": 7.845e-06, + "loss": 0.0685, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.5231291651725769, + "learning_rate": 7.895e-06, + "loss": 0.0717, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.50368332862854, + "learning_rate": 7.945e-06, + "loss": 0.0694, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.4126806855201721, + "learning_rate": 7.995e-06, + "loss": 0.0693, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.719352662563324, + "learning_rate": 8.045e-06, + "loss": 0.0651, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.4405677020549774, + "learning_rate": 8.095e-06, + "loss": 0.0645, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.34746286273002625, + "learning_rate": 8.145e-06, + "loss": 0.0687, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.43200597167015076, + "learning_rate": 8.195e-06, + "loss": 0.0655, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.3464098274707794, + "learning_rate": 8.245e-06, + "loss": 0.0639, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.27263233065605164, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0629, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.343181848526001, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3949901759624481, + "learning_rate": 8.395e-06, + "loss": 0.0643, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.4489574730396271, + "learning_rate": 8.445e-06, + "loss": 0.0698, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.4746992588043213, + "learning_rate": 8.495e-06, + "loss": 0.0702, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.3651655316352844, + "learning_rate": 8.545e-06, + "loss": 0.0673, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.471071720123291, + "learning_rate": 8.595e-06, + "loss": 0.0648, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.35623008012771606, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0698, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.41837432980537415, + "learning_rate": 8.695e-06, + "loss": 0.0639, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.45608940720558167, + "learning_rate": 8.745e-06, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.40346869826316833, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0625, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.479824960231781, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0703, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.38570791482925415, + "learning_rate": 8.895e-06, + "loss": 0.0643, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.36530473828315735, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0652, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37868866324424744, + "learning_rate": 8.995000000000001e-06, + "loss": 0.068, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.32546791434288025, + "learning_rate": 9.045e-06, + "loss": 0.0636, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4664970338344574, + "learning_rate": 9.095e-06, + "loss": 0.0671, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4767569601535797, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0625, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.41823142766952515, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0685, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.3709985613822937, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0681, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.3933296799659729, + "learning_rate": 9.295e-06, + "loss": 0.0678, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.3079732060432434, + "learning_rate": 9.345e-06, + "loss": 0.0598, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.33485978841781616, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0644, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.40312913060188293, + "learning_rate": 9.445e-06, + "loss": 0.0684, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.4334392547607422, + "learning_rate": 9.495e-06, + "loss": 0.0622, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.5787541270256042, + "learning_rate": 9.545e-06, + "loss": 0.0656, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.2827294170856476, + "learning_rate": 9.595e-06, + "loss": 0.0628, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4431678056716919, + "learning_rate": 9.645e-06, + "loss": 0.0633, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.3916102647781372, + "learning_rate": 9.695e-06, + "loss": 0.0632, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.41470348834991455, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0612, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.3224067986011505, + "learning_rate": 9.795e-06, + "loss": 0.061, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.4110186696052551, + "learning_rate": 9.845e-06, + "loss": 0.066, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.4786360561847687, + "learning_rate": 9.895e-06, + "loss": 0.0648, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3667179048061371, + "learning_rate": 9.945e-06, + "loss": 0.0663, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.31722185015678406, + "learning_rate": 9.995e-06, + "loss": 0.0588, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.29271745681762695, + "learning_rate": 1.0045e-05, + "loss": 0.0636, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.3536592423915863, + "learning_rate": 1.0095e-05, + "loss": 0.0625, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.3228510916233063, + "learning_rate": 1.0145e-05, + "loss": 0.067, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.38474392890930176, + "learning_rate": 1.0195e-05, + "loss": 0.0609, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.40410760045051575, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0638, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.3632320463657379, + "learning_rate": 1.0295e-05, + "loss": 0.0587, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.37775692343711853, + "learning_rate": 1.0345e-05, + "loss": 0.0673, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.3343006372451782, + "learning_rate": 1.0395e-05, + "loss": 0.0599, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3232511281967163, + "learning_rate": 1.0445e-05, + "loss": 0.0661, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.38730430603027344, + "learning_rate": 1.0495e-05, + "loss": 0.0649, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.38972100615501404, + "learning_rate": 1.0545e-05, + "loss": 0.0582, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.27242428064346313, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0583, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.30514073371887207, + "learning_rate": 1.0645e-05, + "loss": 0.065, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.45364341139793396, + "learning_rate": 1.0695e-05, + "loss": 0.0576, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.3863564729690552, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0579, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.33717092871665955, + "learning_rate": 1.0795e-05, + "loss": 0.0565, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38073790073394775, + "learning_rate": 1.0845e-05, + "loss": 0.0585, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4277617037296295, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0622, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.3545698821544647, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0609, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.30646640062332153, + "learning_rate": 1.0995e-05, + "loss": 0.0659, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.3607352375984192, + "learning_rate": 1.1045e-05, + "loss": 0.0621, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3624425530433655, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0626, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3856692612171173, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0604, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.4390200972557068, + "learning_rate": 1.1195e-05, + "loss": 0.0601, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.4108872711658478, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0618, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.32732832431793213, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0636, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.3602239787578583, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.29978248476982117, + "learning_rate": 1.1395e-05, + "loss": 0.058, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.48490920662879944, + "learning_rate": 1.1445e-05, + "loss": 0.0614, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.3278113901615143, + "learning_rate": 1.1495e-05, + "loss": 0.0611, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.37862712144851685, + "learning_rate": 1.1545e-05, + "loss": 0.0626, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.3514714241027832, + "learning_rate": 1.1595e-05, + "loss": 0.0583, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.27256545424461365, + "learning_rate": 1.1645e-05, + "loss": 0.062, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.3305833339691162, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0572, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.2881031930446625, + "learning_rate": 1.1745e-05, + "loss": 0.0641, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.3219377398490906, + "learning_rate": 1.1795e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.34189504384994507, + "learning_rate": 1.1845e-05, + "loss": 0.0607, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30541297793388367, + "learning_rate": 1.1895e-05, + "loss": 0.056, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.35898879170417786, + "learning_rate": 1.1945e-05, + "loss": 0.0598, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.33221277594566345, + "learning_rate": 1.1995e-05, + "loss": 0.0603, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.3435590863227844, + "learning_rate": 1.2045e-05, + "loss": 0.0553, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.24820426106452942, + "learning_rate": 1.2095e-05, + "loss": 0.0644, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.3749466836452484, + "learning_rate": 1.2145e-05, + "loss": 0.0615, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.3984697759151459, + "learning_rate": 1.2195e-05, + "loss": 0.0595, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3250048756599426, + "learning_rate": 1.2245e-05, + "loss": 0.0608, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.2794302701950073, + "learning_rate": 1.2295e-05, + "loss": 0.0578, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3811834752559662, + "learning_rate": 1.2345e-05, + "loss": 0.0612, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3160156309604645, + "learning_rate": 1.2395e-05, + "loss": 0.0541, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.3825189173221588, + "learning_rate": 1.2445e-05, + "loss": 0.0621, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.3966415226459503, + "learning_rate": 1.2495e-05, + "loss": 0.0598, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3615199625492096, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0616, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.440762996673584, + "learning_rate": 1.2595e-05, + "loss": 0.0599, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.345599889755249, + "learning_rate": 1.2645e-05, + "loss": 0.0597, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.2963278889656067, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0556, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.3720734417438507, + "learning_rate": 1.2745e-05, + "loss": 0.0618, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3937445878982544, + "learning_rate": 1.2795e-05, + "loss": 0.0624, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.4492146372795105, + "learning_rate": 1.2845e-05, + "loss": 0.0573, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.5172158479690552, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0594, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.3796320855617523, + "learning_rate": 1.2945e-05, + "loss": 0.0528, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.42423033714294434, + "learning_rate": 1.2995e-05, + "loss": 0.0615, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.4578545093536377, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0562, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.38345491886138916, + "learning_rate": 1.3095e-05, + "loss": 0.0574, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.2539816200733185, + "learning_rate": 1.3145e-05, + "loss": 0.0562, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.44323891401290894, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0588, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.4074377417564392, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0591, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.5252887606620789, + "learning_rate": 1.3295e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.39536044001579285, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0588, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.6300645470619202, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0584, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.312466561794281, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0588, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3810383379459381, + "learning_rate": 1.3495e-05, + "loss": 0.0597, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.30645400285720825, + "learning_rate": 1.3545e-05, + "loss": 0.0589, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.2857605218887329, + "learning_rate": 1.3595e-05, + "loss": 0.0589, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.3206622898578644, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0597, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3226073086261749, + "learning_rate": 1.3695e-05, + "loss": 0.0559, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.24912619590759277, + "learning_rate": 1.3745e-05, + "loss": 0.0547, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.29700762033462524, + "learning_rate": 1.3795e-05, + "loss": 0.0563, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.32627588510513306, + "learning_rate": 1.3845e-05, + "loss": 0.0602, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.48569780588150024, + "learning_rate": 1.3895e-05, + "loss": 0.0553, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.323607474565506, + "learning_rate": 1.3945e-05, + "loss": 0.0578, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.42626142501831055, + "learning_rate": 1.3995e-05, + "loss": 0.0539, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.3281888961791992, + "learning_rate": 1.4045e-05, + "loss": 0.0638, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.38216397166252136, + "learning_rate": 1.4095e-05, + "loss": 0.0572, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.2816051244735718, + "learning_rate": 1.4145e-05, + "loss": 0.0543, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.2604160010814667, + "learning_rate": 1.4195e-05, + "loss": 0.0555, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.34318771958351135, + "learning_rate": 1.4245e-05, + "loss": 0.0565, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.31248190999031067, + "learning_rate": 1.4295e-05, + "loss": 0.0583, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.3014426827430725, + "learning_rate": 1.4345e-05, + "loss": 0.0586, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3650360703468323, + "learning_rate": 1.4395e-05, + "loss": 0.0581, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.33519116044044495, + "learning_rate": 1.4445e-05, + "loss": 0.0611, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.3595008850097656, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.054, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.32799217104911804, + "learning_rate": 1.4545e-05, + "loss": 0.0574, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.29410457611083984, + "learning_rate": 1.4595e-05, + "loss": 0.0582, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.2727576792240143, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0588, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.36562401056289673, + "learning_rate": 1.4695e-05, + "loss": 0.0597, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.3351168930530548, + "learning_rate": 1.4745e-05, + "loss": 0.0558, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.23880337178707123, + "learning_rate": 1.4795e-05, + "loss": 0.0571, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.3910059928894043, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0573, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.3103587031364441, + "learning_rate": 1.4895e-05, + "loss": 0.0578, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.3265620768070221, + "learning_rate": 1.4945e-05, + "loss": 0.0542, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.3060731291770935, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0537, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.41589322686195374, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0546, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.3102424740791321, + "learning_rate": 1.5095e-05, + "loss": 0.0564, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.32286810874938965, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0519, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.3129369020462036, + "learning_rate": 1.5195e-05, + "loss": 0.0566, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.43735018372535706, + "learning_rate": 1.5245e-05, + "loss": 0.0552, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.440449595451355, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0589, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.3631616532802582, + "learning_rate": 1.5345e-05, + "loss": 0.0511, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.29124221205711365, + "learning_rate": 1.5395e-05, + "loss": 0.055, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.4247622489929199, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0562, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.36976534128189087, + "learning_rate": 1.5495e-05, + "loss": 0.0604, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.3580514192581177, + "learning_rate": 1.5545e-05, + "loss": 0.0547, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.2969493567943573, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0564, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28583595156669617, + "learning_rate": 1.5645e-05, + "loss": 0.0586, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.3220328390598297, + "learning_rate": 1.5695e-05, + "loss": 0.0588, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.32270652055740356, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0559, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.3170712888240814, + "learning_rate": 1.5795e-05, + "loss": 0.0566, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.2966325879096985, + "learning_rate": 1.5845e-05, + "loss": 0.0551, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.2788766622543335, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.051, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.3521708548069, + "learning_rate": 1.5945e-05, + "loss": 0.0558, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.35105040669441223, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0551, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.4221852123737335, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0573, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.4217511713504791, + "learning_rate": 1.6095e-05, + "loss": 0.0532, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.4519175589084625, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0489, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3427059054374695, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0544, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29204127192497253, + "learning_rate": 1.6245e-05, + "loss": 0.0525, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.24138635396957397, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0559, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.3599991500377655, + "learning_rate": 1.6345e-05, + "loss": 0.0544, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.3326258361339569, + "learning_rate": 1.6395e-05, + "loss": 0.0593, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31976261734962463, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0624, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.26379555463790894, + "learning_rate": 1.6495e-05, + "loss": 0.0529, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2576923966407776, + "learning_rate": 1.6545e-05, + "loss": 0.0527, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.30890268087387085, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0571, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.4547998309135437, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.3050350248813629, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0552, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.2984176278114319, + "learning_rate": 1.6745e-05, + "loss": 0.0528, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.34583792090415955, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0512, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.33085328340530396, + "learning_rate": 1.6845e-05, + "loss": 0.0563, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.33391326665878296, + "learning_rate": 1.6895e-05, + "loss": 0.0527, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.23657914996147156, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0509, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.3719659149646759, + "learning_rate": 1.6995e-05, + "loss": 0.0568, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.4268389940261841, + "learning_rate": 1.7045e-05, + "loss": 0.0525, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.2726958394050598, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0482, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27847009897232056, + "learning_rate": 1.7145e-05, + "loss": 0.0559, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.7195e-05, + "loss": 0.0588, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.2666028141975403, + "learning_rate": 1.7245e-05, + "loss": 0.0548, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.26048436760902405, + "learning_rate": 1.7295e-05, + "loss": 0.0592, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.22569318115711212, + "learning_rate": 1.7345e-05, + "loss": 0.0505, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.35350096225738525, + "learning_rate": 1.7395e-05, + "loss": 0.0566, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.3166314363479614, + "learning_rate": 1.7445e-05, + "loss": 0.0517, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.24918541312217712, + "learning_rate": 1.7495e-05, + "loss": 0.0545, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3416207730770111, + "learning_rate": 1.7545e-05, + "loss": 0.0544, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.22633257508277893, + "learning_rate": 1.7595e-05, + "loss": 0.0545, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.3972736597061157, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0591, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3213656544685364, + "learning_rate": 1.7695e-05, + "loss": 0.0537, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.30361926555633545, + "learning_rate": 1.7745e-05, + "loss": 0.0529, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.35521000623703003, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0549, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.4444983899593353, + "learning_rate": 1.7845e-05, + "loss": 0.0529, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.32495826482772827, + "learning_rate": 1.7895e-05, + "loss": 0.0557, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.2893896698951721, + "learning_rate": 1.7945e-05, + "loss": 0.056, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.2607886493206024, + "learning_rate": 1.7995e-05, + "loss": 0.055, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.3305937647819519, + "learning_rate": 1.8045e-05, + "loss": 0.0529, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.3003319799900055, + "learning_rate": 1.8095e-05, + "loss": 0.0515, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3523712456226349, + "learning_rate": 1.8145e-05, + "loss": 0.0524, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.27575406432151794, + "learning_rate": 1.8195e-05, + "loss": 0.0535, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.36445286870002747, + "learning_rate": 1.8245e-05, + "loss": 0.0552, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.41415491700172424, + "learning_rate": 1.8295e-05, + "loss": 0.0537, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.3738953471183777, + "learning_rate": 1.8345e-05, + "loss": 0.0555, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.3446105420589447, + "learning_rate": 1.8395e-05, + "loss": 0.0526, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.39016082882881165, + "learning_rate": 1.8445e-05, + "loss": 0.0536, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.4075721502304077, + "learning_rate": 1.8495e-05, + "loss": 0.0548, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.4071345925331116, + "learning_rate": 1.8545e-05, + "loss": 0.0539, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.34446755051612854, + "learning_rate": 1.8595e-05, + "loss": 0.057, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.3431609272956848, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0515, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.31537994742393494, + "learning_rate": 1.8695e-05, + "loss": 0.0534, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.4846970736980438, + "learning_rate": 1.8745e-05, + "loss": 0.0532, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.3109683394432068, + "learning_rate": 1.8795e-05, + "loss": 0.0547, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.28903284668922424, + "learning_rate": 1.8845e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.26592350006103516, + "learning_rate": 1.8895e-05, + "loss": 0.0499, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.27347302436828613, + "learning_rate": 1.8945e-05, + "loss": 0.052, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.3194405138492584, + "learning_rate": 1.8995e-05, + "loss": 0.0521, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.40457430481910706, + "learning_rate": 1.9045e-05, + "loss": 0.0523, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.3055919408798218, + "learning_rate": 1.9095e-05, + "loss": 0.0504, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.4911873936653137, + "learning_rate": 1.9145e-05, + "loss": 0.0531, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.30619072914123535, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0525, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.24404728412628174, + "learning_rate": 1.9245e-05, + "loss": 0.0498, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.31723177433013916, + "learning_rate": 1.9295e-05, + "loss": 0.0519, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.28105342388153076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0528, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.32965341210365295, + "learning_rate": 1.9395e-05, + "loss": 0.0504, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.2487417608499527, + "learning_rate": 1.9445e-05, + "loss": 0.0552, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3788810670375824, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0535, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.2561226189136505, + "learning_rate": 1.9545e-05, + "loss": 0.056, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.29964715242385864, + "learning_rate": 1.9595e-05, + "loss": 0.0565, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.386600136756897, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0468, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.37690475583076477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.35367539525032043, + "learning_rate": 1.9745e-05, + "loss": 0.0501, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.30189844965934753, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0524, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.27525416016578674, + "learning_rate": 1.9845e-05, + "loss": 0.0524, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.22557462751865387, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.3572654128074646, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0536, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.37130483984947205, + "learning_rate": 1.9995e-05, + "loss": 0.0507, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.40820422768592834, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0547, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.24222403764724731, + "learning_rate": 2.0095e-05, + "loss": 0.0536, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.35473307967185974, + "learning_rate": 2.0145e-05, + "loss": 0.0503, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24238434433937073, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0511, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.32721346616744995, + "learning_rate": 2.0245e-05, + "loss": 0.0474, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.3787407875061035, + "learning_rate": 2.0295e-05, + "loss": 0.0515, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.3435424566268921, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0521, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.250515341758728, + "learning_rate": 2.0395e-05, + "loss": 0.0539, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.2730806767940521, + "learning_rate": 2.0445e-05, + "loss": 0.0544, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.32704228162765503, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0516, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.26865822076797485, + "learning_rate": 2.0545e-05, + "loss": 0.0557, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.31514087319374084, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.051, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.2616795301437378, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0483, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2979176938533783, + "learning_rate": 2.0695e-05, + "loss": 0.056, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.2876877188682556, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0496, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.2801620662212372, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0533, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.3521801233291626, + "learning_rate": 2.0845e-05, + "loss": 0.051, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.3016152083873749, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0515, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.29837727546691895, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0483, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.255982369184494, + "learning_rate": 2.0995e-05, + "loss": 0.0482, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.20852255821228027, + "learning_rate": 2.1045e-05, + "loss": 0.05, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.2746782898902893, + "learning_rate": 2.1095e-05, + "loss": 0.0513, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.28745001554489136, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0486, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.3163246810436249, + "learning_rate": 2.1195e-05, + "loss": 0.0536, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.2279432713985443, + "learning_rate": 2.1245e-05, + "loss": 0.0497, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.2273164689540863, + "learning_rate": 2.1295e-05, + "loss": 0.0485, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.3124007284641266, + "learning_rate": 2.1345e-05, + "loss": 0.0473, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.2097717970609665, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0514, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.24428582191467285, + "learning_rate": 2.1445e-05, + "loss": 0.0543, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.47236987948417664, + "learning_rate": 2.1495e-05, + "loss": 0.0525, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.3367164433002472, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0525, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.34086376428604126, + "learning_rate": 2.1595e-05, + "loss": 0.049, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.3060646951198578, + "learning_rate": 2.1645e-05, + "loss": 0.0538, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.2997030019760132, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0553, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.2652549743652344, + "learning_rate": 2.1745e-05, + "loss": 0.051, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.3072361350059509, + "learning_rate": 2.1795e-05, + "loss": 0.0516, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2494215965270996, + "learning_rate": 2.1845e-05, + "loss": 0.0476, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.20796196162700653, + "learning_rate": 2.1895e-05, + "loss": 0.0553, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.24235095083713531, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.21033242344856262, + "learning_rate": 2.1995e-05, + "loss": 0.0542, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.26056164503097534, + "learning_rate": 2.2045e-05, + "loss": 0.0546, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.24798662960529327, + "learning_rate": 2.2095e-05, + "loss": 0.0529, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.22179707884788513, + "learning_rate": 2.2145e-05, + "loss": 0.0542, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.2660025358200073, + "learning_rate": 2.2195e-05, + "loss": 0.0496, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.3771500289440155, + "learning_rate": 2.2245e-05, + "loss": 0.0502, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.3059975206851959, + "learning_rate": 2.2295e-05, + "loss": 0.0482, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.2941029667854309, + "learning_rate": 2.2345e-05, + "loss": 0.0483, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3481833040714264, + "learning_rate": 2.2395e-05, + "loss": 0.0541, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.2729704678058624, + "learning_rate": 2.2445e-05, + "loss": 0.0514, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.44202038645744324, + "learning_rate": 2.2495e-05, + "loss": 0.0534, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.311088889837265, + "learning_rate": 2.2545e-05, + "loss": 0.0474, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30663055181503296, + "learning_rate": 2.2595e-05, + "loss": 0.0491, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.3530942499637604, + "learning_rate": 2.2645e-05, + "loss": 0.0503, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.27998727560043335, + "learning_rate": 2.2695e-05, + "loss": 0.0493, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.2679221034049988, + "learning_rate": 2.2745e-05, + "loss": 0.0512, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.348821222782135, + "learning_rate": 2.2795e-05, + "loss": 0.0489, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.260958194732666, + "learning_rate": 2.2845e-05, + "loss": 0.0474, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.2429468184709549, + "learning_rate": 2.2895e-05, + "loss": 0.053, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.2635575234889984, + "learning_rate": 2.2945e-05, + "loss": 0.053, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.3416514992713928, + "learning_rate": 2.2995e-05, + "loss": 0.0476, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.221172496676445, + "learning_rate": 2.3045e-05, + "loss": 0.0519, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.30541878938674927, + "learning_rate": 2.3095e-05, + "loss": 0.0522, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.25274574756622314, + "learning_rate": 2.3145e-05, + "loss": 0.0499, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.366253525018692, + "learning_rate": 2.3195e-05, + "loss": 0.0463, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.16825024783611298, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0493, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.3158467710018158, + "learning_rate": 2.3295e-05, + "loss": 0.0482, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.2887287139892578, + "learning_rate": 2.3345e-05, + "loss": 0.053, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.4717058539390564, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0515, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.3148113489151001, + "learning_rate": 2.3445e-05, + "loss": 0.0523, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.23815515637397766, + "learning_rate": 2.3495e-05, + "loss": 0.0523, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23897255957126617, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0484, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.2680213749408722, + "learning_rate": 2.3595e-05, + "loss": 0.0485, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.29999756813049316, + "learning_rate": 2.3645e-05, + "loss": 0.0506, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.30607151985168457, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.05, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.26077136397361755, + "learning_rate": 2.3745e-05, + "loss": 0.0558, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2638765275478363, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.049, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.24569877982139587, + "learning_rate": 2.3845e-05, + "loss": 0.053, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.29622989892959595, + "learning_rate": 2.3895e-05, + "loss": 0.0517, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.2598831355571747, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.05, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.25716227293014526, + "learning_rate": 2.3995e-05, + "loss": 0.0453, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.37305381894111633, + "learning_rate": 2.4045e-05, + "loss": 0.0539, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.3151630461215973, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0516, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.296941339969635, + "learning_rate": 2.4145e-05, + "loss": 0.0509, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.31995704770088196, + "learning_rate": 2.4195e-05, + "loss": 0.0498, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2426208257675171, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0475, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.23927678167819977, + "learning_rate": 2.4295e-05, + "loss": 0.05, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.1887904703617096, + "learning_rate": 2.4345e-05, + "loss": 0.05, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.18480589985847473, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0496, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.24515867233276367, + "learning_rate": 2.4445e-05, + "loss": 0.0464, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.25505101680755615, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0481, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.24435125291347504, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0502, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.25899314880371094, + "learning_rate": 2.4595e-05, + "loss": 0.0535, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.30518922209739685, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0481, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.26098036766052246, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0477, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.24723340570926666, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.26713746786117554, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0506, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.21166373789310455, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0502, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2602159082889557, + "learning_rate": 2.4895e-05, + "loss": 0.0506, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27918240427970886, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.049, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.2131885290145874, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0477, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.28778040409088135, + "learning_rate": 2.5045e-05, + "loss": 0.0487, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.3056010603904724, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.049, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.2850024700164795, + "learning_rate": 2.5145e-05, + "loss": 0.0475, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.20490600168704987, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0497, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.32371339201927185, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0491, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.32466208934783936, + "learning_rate": 2.5295e-05, + "loss": 0.0485, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.2779853343963623, + "learning_rate": 2.5345e-05, + "loss": 0.0485, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.3257012367248535, + "learning_rate": 2.5395e-05, + "loss": 0.0486, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2625649571418762, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0486, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.19846542179584503, + "learning_rate": 2.5495e-05, + "loss": 0.0494, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.2415613830089569, + "learning_rate": 2.5545e-05, + "loss": 0.0539, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.20438073575496674, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0476, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.2543872892856598, + "learning_rate": 2.5645e-05, + "loss": 0.0489, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.26969635486602783, + "learning_rate": 2.5695e-05, + "loss": 0.0557, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.27818888425827026, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0517, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.28826993703842163, + "learning_rate": 2.5795e-05, + "loss": 0.0499, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.19518250226974487, + "learning_rate": 2.5845e-05, + "loss": 0.0499, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.2639068067073822, + "learning_rate": 2.5895e-05, + "loss": 0.0459, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.26014742255210876, + "learning_rate": 2.5945e-05, + "loss": 0.0522, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.26659756898880005, + "learning_rate": 2.5995e-05, + "loss": 0.0471, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.24073593318462372, + "learning_rate": 2.6045e-05, + "loss": 0.0488, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.2724616527557373, + "learning_rate": 2.6095e-05, + "loss": 0.0493, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.3012109398841858, + "learning_rate": 2.6145e-05, + "loss": 0.049, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23306235671043396, + "learning_rate": 2.6195e-05, + "loss": 0.0474, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.23879121243953705, + "learning_rate": 2.6245e-05, + "loss": 0.048, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20701922476291656, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0527, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.28734415769577026, + "learning_rate": 2.6345e-05, + "loss": 0.0469, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.26227426528930664, + "learning_rate": 2.6395e-05, + "loss": 0.0472, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.24196277558803558, + "learning_rate": 2.6445e-05, + "loss": 0.0462, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.3405964970588684, + "learning_rate": 2.6495e-05, + "loss": 0.0479, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.24121852219104767, + "learning_rate": 2.6545e-05, + "loss": 0.0444, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.284856379032135, + "learning_rate": 2.6595e-05, + "loss": 0.0492, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.2515311539173126, + "learning_rate": 2.6645e-05, + "loss": 0.0453, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.29297393560409546, + "learning_rate": 2.6695e-05, + "loss": 0.049, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.2528100311756134, + "learning_rate": 2.6745e-05, + "loss": 0.0477, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.29126009345054626, + "learning_rate": 2.6795e-05, + "loss": 0.0484, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.26489511132240295, + "learning_rate": 2.6845e-05, + "loss": 0.0476, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.33150023221969604, + "learning_rate": 2.6895e-05, + "loss": 0.0475, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.25904157757759094, + "learning_rate": 2.6945e-05, + "loss": 0.0495, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.24830900132656097, + "learning_rate": 2.6995e-05, + "loss": 0.0476, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.277997225522995, + "learning_rate": 2.7045e-05, + "loss": 0.0478, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.3100328743457794, + "learning_rate": 2.7095e-05, + "loss": 0.0501, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.2957077622413635, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0491, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.19402435421943665, + "learning_rate": 2.7195e-05, + "loss": 0.0426, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.20080482959747314, + "learning_rate": 2.7245e-05, + "loss": 0.0464, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.21148671209812164, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0457, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.2299162596464157, + "learning_rate": 2.7345e-05, + "loss": 0.0467, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2296467125415802, + "learning_rate": 2.7395e-05, + "loss": 0.0489, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.25099772214889526, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0487, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2503034472465515, + "learning_rate": 2.7495e-05, + "loss": 0.0533, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.24678738415241241, + "learning_rate": 2.7545e-05, + "loss": 0.0462, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.3281477987766266, + "learning_rate": 2.7595e-05, + "loss": 0.0492, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.32214412093162537, + "learning_rate": 2.7645e-05, + "loss": 0.0459, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.29244086146354675, + "learning_rate": 2.7695e-05, + "loss": 0.0495, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.2372455894947052, + "learning_rate": 2.7745e-05, + "loss": 0.0477, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.28411489725112915, + "learning_rate": 2.7795e-05, + "loss": 0.0514, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.36217001080513, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2913915514945984, + "learning_rate": 2.7895e-05, + "loss": 0.0535, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.3237503170967102, + "learning_rate": 2.7945e-05, + "loss": 0.0509, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.32016751170158386, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0526, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.25484171509742737, + "learning_rate": 2.8045e-05, + "loss": 0.0508, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.22641469538211823, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.23269280791282654, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0484, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23346202075481415, + "learning_rate": 2.8195e-05, + "loss": 0.0472, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.25522616505622864, + "learning_rate": 2.8245e-05, + "loss": 0.0493, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.2478165179491043, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0468, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.25732970237731934, + "learning_rate": 2.8345e-05, + "loss": 0.0484, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.21047812700271606, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0455, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.252676397562027, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0496, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20048962533473969, + "learning_rate": 2.8495e-05, + "loss": 0.0434, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.1821785420179367, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0502, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.24106337130069733, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0471, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.413974404335022, + "learning_rate": 2.8645e-05, + "loss": 0.0542, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.29419073462486267, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0499, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.20694036781787872, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0453, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2390667200088501, + "learning_rate": 2.8795e-05, + "loss": 0.045, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2563699781894684, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.046, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.22678422927856445, + "learning_rate": 2.8895e-05, + "loss": 0.0456, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.26104822754859924, + "learning_rate": 2.8945e-05, + "loss": 0.0446, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.22175569832324982, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0478, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.20990405976772308, + "learning_rate": 2.9045e-05, + "loss": 0.0459, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.27277135848999023, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0477, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.22422750294208527, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.047, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.2740926742553711, + "learning_rate": 2.9195e-05, + "loss": 0.0444, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2977927327156067, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0477, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26431694626808167, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0488, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.22759316861629486, + "learning_rate": 2.9345e-05, + "loss": 0.0465, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.24671563506126404, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0501, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.24065586924552917, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0439, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.3334764838218689, + "learning_rate": 2.9495e-05, + "loss": 0.0475, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.23565605282783508, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.046, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.26102808117866516, + "learning_rate": 2.9595e-05, + "loss": 0.0494, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.34544840455055237, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.05, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.2476256936788559, + "learning_rate": 2.9695e-05, + "loss": 0.0477, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.2795628309249878, + "learning_rate": 2.9745e-05, + "loss": 0.0508, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.2756715714931488, + "learning_rate": 2.9795e-05, + "loss": 0.0484, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.20578329265117645, + "learning_rate": 2.9845e-05, + "loss": 0.0461, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.20354008674621582, + "learning_rate": 2.9895e-05, + "loss": 0.0431, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24710936844348907, + "learning_rate": 2.9945e-05, + "loss": 0.0459, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.32639119029045105, + "learning_rate": 2.9995e-05, + "loss": 0.045, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.27762553095817566, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0476, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.2903423607349396, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.3292105793952942, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0503, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.23503656685352325, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0446, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.24153585731983185, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.048, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.24157844483852386, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.046, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.23378467559814453, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0478, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.20753100514411926, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0527, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.2574090361595154, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0468, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.23341813683509827, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0445, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21378666162490845, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0487, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.2410328984260559, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0488, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.22005778551101685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0478, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.3197212219238281, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.046, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.253745436668396, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0437, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.1991635262966156, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0474, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.20773734152317047, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0439, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.3026197552680969, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0462, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.22027666866779327, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0443, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.2512618899345398, + "learning_rate": 2.99997744429684e-05, + "loss": 0.044, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.23001381754875183, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0479, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.20057332515716553, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0433, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.2831740975379944, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0462, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.18362601101398468, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0472, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.29741403460502625, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0443, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.20039376616477966, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0457, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.14999084174633026, + "learning_rate": 2.999958785135414e-05, + "loss": 0.043, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.18134453892707825, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0416, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.25010257959365845, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0494, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.31700536608695984, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0455, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.25080761313438416, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0436, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.2008715718984604, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0459, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.21441133320331573, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0492, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.23232029378414154, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0459, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.30173608660697937, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0505, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.21179363131523132, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0438, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.21905283629894257, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0439, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.17932383716106415, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0452, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.21132917702198029, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0454, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.22732721269130707, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.047, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.28214913606643677, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0446, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.23979637026786804, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0454, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.3095552921295166, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0449, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.2585015594959259, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0473, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.18226519227027893, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0472, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2568792402744293, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0443, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.27914026379585266, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0423, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.21452564001083374, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0457, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.22575126588344574, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0458, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.24053922295570374, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0453, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.21956735849380493, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.048, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2325354665517807, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0439, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.1692194789648056, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.044, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.25796884298324585, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0416, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.2719244956970215, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0445, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.23312833905220032, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0488, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.24458816647529602, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0473, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.24380528926849365, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0411, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.1836114227771759, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0433, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.21051307022571564, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0462, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.18093563616275787, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.045, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2444952130317688, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0485, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.19429947435855865, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0487, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.2294744998216629, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0479, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.2147439569234848, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0492, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.2777695059776306, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0419, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22944948077201843, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0458, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.2822294235229492, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0429, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.20201396942138672, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0435, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.1961182802915573, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0441, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.1653209924697876, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0473, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.2082156240940094, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0508, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.199684277176857, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0466, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.20871016383171082, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0436, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.22781118750572205, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0504, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2652275562286377, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0467, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.20812813937664032, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0456, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2533996105194092, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0458, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.18139654397964478, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0441, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.27243369817733765, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0472, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.326709121465683, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0487, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.26698657870292664, + "learning_rate": 2.999617967180403e-05, + "loss": 0.05, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.4119804799556732, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0455, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.21175070106983185, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0462, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.22503416240215302, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.0468, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.254734605550766, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0471, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26713427901268005, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0439, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.2739362120628357, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0455, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.26221534609794617, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0451, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26752719283103943, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.23052464425563812, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0481, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.162479966878891, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0435, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.24716691672801971, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0459, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.14981329441070557, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0407, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.23326952755451202, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0459, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.2200203239917755, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.046, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.21501530706882477, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0433, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.22950918972492218, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0454, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.2042679488658905, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0444, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.16468611359596252, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0434, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.2180141657590866, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0447, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.1658746749162674, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.24921023845672607, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0398, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0445, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.25189560651779175, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0429, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.2614912688732147, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0427, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.2890806794166565, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0458, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.24215207993984222, + "learning_rate": 2.999336925698801e-05, + "loss": 0.04, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.1806204915046692, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0435, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.18884161114692688, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0422, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.2760990262031555, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0427, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.22660709917545319, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0459, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.24862384796142578, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.045, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.1911306381225586, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0433, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.2612020969390869, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0414, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.20945195853710175, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.043, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.2436099350452423, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0472, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.23788926005363464, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0455, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.20128341019153595, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0456, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.24002164602279663, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0414, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.2653037905693054, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.042, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.15122903883457184, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0395, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.19299687445163727, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0452, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.2534617781639099, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0445, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.29057401418685913, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0398, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.19744732975959778, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.041, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.25267109274864197, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0439, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2406993955373764, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0445, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.2561663091182709, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0465, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.2137032002210617, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0407, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.3002893030643463, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0433, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.32190272212028503, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0444, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.2076684534549713, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0432, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.31036099791526794, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0431, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.21386218070983887, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0442, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.19012001156806946, + "learning_rate": 2.998948184763285e-05, + "loss": 0.048, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.2310018688440323, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0466, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.1984279453754425, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0443, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.24101053178310394, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0446, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.17791084945201874, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0403, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.2016495168209076, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0404, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.19435155391693115, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.21197110414505005, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0398, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.22241441905498505, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.20825718343257904, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0434, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.24533888697624207, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.046, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.2455352544784546, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0412, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.18937505781650543, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0425, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.22356615960597992, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0418, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.23797748982906342, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0457, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.23130296170711517, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0427, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.20548853278160095, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0439, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19323743879795074, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0432, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.24339382350444794, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0412, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.18780124187469482, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0416, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.19353246688842773, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0461, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.21522872149944305, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0412, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.26847484707832336, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.3781932294368744, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0401, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.3219687342643738, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0423, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.1699237823486328, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0421, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.2174658179283142, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0398, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.2433294802904129, + "learning_rate": 2.998488804893431e-05, + "loss": 0.042, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.2116415947675705, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0405, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.22813065350055695, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0432, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.24252736568450928, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.048, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.2009168565273285, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0431, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.19379755854606628, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.23313777148723602, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0384, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.22056345641613007, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0459, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.3403800129890442, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0439, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.17649835348129272, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0419, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.20594412088394165, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0418, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.2901265323162079, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0405, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.22897225618362427, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0424, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.20961393415927887, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0438, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.28585514426231384, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0477, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2239968478679657, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0398, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.1651894897222519, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0405, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.1731943041086197, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0407, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.1708967089653015, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0417, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.17396977543830872, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0359, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.21327704191207886, + "learning_rate": 2.998095036004704e-05, + "loss": 0.045, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.15585461258888245, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0403, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.21774765849113464, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0403, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.1889292299747467, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0408, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.1639491617679596, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0425, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.23745876550674438, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.043, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19481611251831055, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0453, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.21001650393009186, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0429, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.2114192098379135, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0419, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.2506379783153534, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0392, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.21439281105995178, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0422, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.24763159453868866, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0433, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.1719428300857544, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0422, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.25943970680236816, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0425, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22226977348327637, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0448, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.2018195390701294, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0408, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.20869500935077667, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0443, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.23311494290828705, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.042, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.2501373589038849, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0382, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.2650911509990692, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0377, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.21386376023292542, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0405, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.27908873558044434, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0431, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.18731699883937836, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0402, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.2566622197628021, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0419, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.2418505847454071, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0406, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.199744313955307, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0424, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.20134931802749634, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.042, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.15607263147830963, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0453, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.22614715993404388, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0413, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.22810052335262299, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0433, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.16557516157627106, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0424, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.20254060626029968, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0394, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2068638950586319, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0408, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.27261367440223694, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0433, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.23216570913791656, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0445, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.2546656131744385, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0381, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.21829853951931, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0472, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.17109332978725433, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0396, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.1746513843536377, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0413, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.22411789000034332, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0373, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.21182285249233246, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.038, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.19430142641067505, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0419, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.22196920216083527, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0395, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.20695635676383972, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0407, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.1721169799566269, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0396, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.23676829040050507, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0396, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.3019765317440033, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0443, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.23958736658096313, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0402, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.17959944903850555, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0434, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.1563226878643036, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0397, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.16746941208839417, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0402, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.26673364639282227, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0411, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.23634585738182068, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0427, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.19724313914775848, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0408, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.2072090059518814, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0389, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.18574072420597076, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0427, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.23390066623687744, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0411, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.2137259542942047, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0395, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.21458280086517334, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0422, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.1838904619216919, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.043, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.1834891140460968, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0389, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.2012249082326889, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.27410435676574707, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0473, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.21607685089111328, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0405, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.17525449395179749, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.043, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.23137639462947845, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0414, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.20958152413368225, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0404, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.21453802287578583, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.043, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.2522047758102417, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.04, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.23595386743545532, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0396, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.23375281691551208, + "learning_rate": 2.996358559911323e-05, + "loss": 0.042, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.21628527343273163, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0434, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.2463727742433548, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0391, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.19919127225875854, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.043, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.28369203209877014, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0463, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.30222833156585693, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0397, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.19382458925247192, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0402, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2075910121202469, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0392, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.2503066658973694, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0427, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.18662676215171814, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0397, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.2432677149772644, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0425, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.19900213181972504, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0422, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.12742096185684204, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.20504114031791687, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0404, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.19127227365970612, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0425, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.18699967861175537, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0419, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.20074906945228577, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0381, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.24741096794605255, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0389, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.22948721051216125, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0435, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.17772004008293152, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0412, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.228739395737648, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0406, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.18312685191631317, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0401, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.21340692043304443, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0403, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.17110474407672882, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0388, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.19640640914440155, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.2386539727449417, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.043, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.20800922811031342, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0412, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1889747679233551, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0377, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.1801699548959732, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0402, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.20975525677204132, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0426, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.22626695036888123, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0404, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.2192756086587906, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0438, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.21307197213172913, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0447, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.28529244661331177, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0412, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.26367515325546265, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.044, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.29029133915901184, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.14019586145877838, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0427, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19176369905471802, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0408, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.19753597676753998, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0395, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1900634765625, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0427, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.21698379516601562, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0452, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.21722441911697388, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0417, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.1760135143995285, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0405, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18030588328838348, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0425, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.20307210087776184, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0416, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.19713939726352692, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0401, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.18719582259655, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.19560874998569489, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.04, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1633717119693756, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0424, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.20183435082435608, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0397, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.23784901201725006, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0438, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.17358708381652832, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0393, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18991878628730774, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0387, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.20455898344516754, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.042, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.23264537751674652, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0447, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2586922347545624, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0402, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.20447152853012085, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0406, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.1814388930797577, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0412, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2679784595966339, + "learning_rate": 2.994497917870469e-05, + "loss": 0.039, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.1818714141845703, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0423, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.2153957635164261, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0391, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.1858970820903778, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0394, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.23248426616191864, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.17587599158287048, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0367, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.18720309436321259, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.039, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.17602671682834625, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0393, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.15901489555835724, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0418, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.18207870423793793, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0401, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.22800953686237335, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0416, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.20031091570854187, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0417, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.24054989218711853, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0438, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.201461061835289, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0369, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.15697605907917023, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0397, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.16894802451133728, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0403, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2567543089389801, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0421, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.20297718048095703, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0396, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.17331846058368683, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.038, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.2214743047952652, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0405, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.2107590287923813, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0388, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.1768934726715088, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0364, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.16018562018871307, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0405, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.18402984738349915, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0425, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.21482129395008087, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0399, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.20554907619953156, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0388, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.2401154637336731, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0369, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.15072989463806152, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0359, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.1649719625711441, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0399, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.2386736422777176, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0403, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.2389005720615387, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0416, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.14930105209350586, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0372, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.18874907493591309, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0356, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.21466751396656036, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0391, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1697709560394287, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.24686385691165924, + "learning_rate": 2.993190369330068e-05, + "loss": 0.04, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.19733406603336334, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0417, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.3981468677520752, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0375, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.20926636457443237, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0384, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.18989330530166626, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0365, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17171184718608856, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0386, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.20474061369895935, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0413, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.19314105808734894, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0358, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.1794378012418747, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0409, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.17263059318065643, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0385, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.17863675951957703, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0355, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.20071984827518463, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0386, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.2025180608034134, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0409, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.21840184926986694, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0397, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.197452113032341, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0379, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.15370187163352966, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0379, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.22221392393112183, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0396, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.18536874651908875, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.18744619190692902, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0378, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.21197383105754852, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0399, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.29221004247665405, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0355, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.2411249279975891, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.039, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.3187369108200073, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.036, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.18096214532852173, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0372, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.2743191719055176, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0401, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.1864418387413025, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0384, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.15518276393413544, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.039, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.18089881539344788, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0359, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.2218618243932724, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0427, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.17051172256469727, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0401, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.15884852409362793, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0378, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1561235636472702, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0383, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.19711007177829742, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0409, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.18595683574676514, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.043, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.18613699078559875, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0357, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.1390007734298706, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0356, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.25603482127189636, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0377, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.1902502328157425, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0381, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.18781402707099915, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0393, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.20504912734031677, + "learning_rate": 2.991569760600352e-05, + "loss": 0.042, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.17359918355941772, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0373, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.2029414027929306, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0412, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.23485742509365082, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0398, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1454097479581833, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0394, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.22343166172504425, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0386, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.17754614353179932, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0358, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.14276635646820068, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0405, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.2062550187110901, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0429, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2093406766653061, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.038, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.14510402083396912, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0374, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.22726939618587494, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0373, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.14905233681201935, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0392, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.15828309953212738, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0361, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.990945968675917e-05, + "loss": 0.037, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.1867590844631195, + "learning_rate": 2.990900562792633e-05, + "loss": 0.038, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.18278446793556213, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0358, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.24628379940986633, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0393, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.2195759266614914, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0372, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.2397889941930771, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0352, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.1563243269920349, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0361, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.16278347373008728, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0366, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1776551604270935, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0369, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.19893717765808105, + "learning_rate": 2.990533239951709e-05, + "loss": 0.039, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.21403104066848755, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0378, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.19885864853858948, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0393, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.18425336480140686, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0404, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.1930677592754364, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0369, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18271560966968536, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0351, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.16518554091453552, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0367, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.17990152537822723, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0349, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.19625109434127808, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0368, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.1891077160835266, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0407, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.20162540674209595, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.17586444318294525, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0346, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.13146457076072693, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0338, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.16059468686580658, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0366, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.1724517047405243, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0395, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.1847251057624817, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0365, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.20104378461837769, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.037, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.2712373733520508, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0347, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.18014875054359436, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0388, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.17479945719242096, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0388, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.24176514148712158, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0387, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.2453809529542923, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0381, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2105131894350052, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0369, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1763688176870346, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0384, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.16223029792308807, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0348, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.173785999417305, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.038, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.23269419372081757, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0393, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.19176329672336578, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0399, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.1559731364250183, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0375, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.15701691806316376, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.22872254252433777, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0371, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15535064041614532, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0378, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.21613062918186188, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0382, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.1504523903131485, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0386, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.177171990275383, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2307550013065338, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0354, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.2215310037136078, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0371, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.23132716119289398, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0389, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.1937142014503479, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0373, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.23848143219947815, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0355, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.23107734322547913, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0359, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1455046832561493, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0404, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.21890339255332947, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0421, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.1897345334291458, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0357, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.2016463577747345, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0363, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.15850050747394562, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0391, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.18169008195400238, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0377, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.26095741987228394, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0391, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.20983633399009705, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0355, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.21026623249053955, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.2104070633649826, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0343, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1801508367061615, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0369, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.22245049476623535, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.0358, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.19220919907093048, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0354, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.2285826951265335, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0391, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.15382888913154602, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0374, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.13551367819309235, + "learning_rate": 2.987759237035982e-05, + "loss": 0.039, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.23362696170806885, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0367, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.15972478687763214, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0319, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.23858226835727692, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0354, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.17988596856594086, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0391, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.21262454986572266, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0392, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.1443355828523636, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0384, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.24133095145225525, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0382, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.175334632396698, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.039, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.20340824127197266, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0376, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.16549751162528992, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0371, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.16527467966079712, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0373, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1418737769126892, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0378, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18754976987838745, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0355, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.13649867475032806, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0329, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.21408453583717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0348, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.20929333567619324, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.24083013832569122, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0378, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17924576997756958, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0369, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.22360439598560333, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0382, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.15963956713676453, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0361, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.19003787636756897, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0365, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2458433210849762, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0375, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.22048939764499664, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0382, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.19965334236621857, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0371, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.228081613779068, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0375, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.24077889323234558, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0347, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.17692111432552338, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0365, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.1996857225894928, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0358, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.24799935519695282, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0377, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.24300269782543182, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0362, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.2230309098958969, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0364, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.16640354692935944, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0346, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.14670157432556152, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.039, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18531887233257294, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0353, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.19566190242767334, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0391, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.1967019885778427, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0355, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.2161232978105545, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0375, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.17630350589752197, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0397, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18905608355998993, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0347, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.1890578716993332, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.036, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.15397298336029053, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0401, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0362, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.20983757078647614, + "learning_rate": 2.985388676461079e-05, + "loss": 0.036, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.1554982215166092, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0378, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.1961488276720047, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0356, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.17265456914901733, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0357, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.1275556981563568, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0353, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.1667783409357071, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0354, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.27920690178871155, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0385, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.12451875954866409, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0352, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.15794622898101807, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0339, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.16285695135593414, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0369, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.18990163505077362, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0376, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.1867877095937729, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0363, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.15270273387432098, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0394, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.1934751272201538, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0382, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16943520307540894, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0369, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.20546840131282806, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0336, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.18490619957447052, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0415, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.16439616680145264, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0384, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.13623777031898499, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0379, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.18209043145179749, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0357, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.19922122359275818, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0365, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.15637168288230896, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0323, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1781233251094818, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0329, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.17547845840454102, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0344, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.1770639270544052, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0358, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.2677673101425171, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0356, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.16740617156028748, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0344, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.13076575100421906, + "learning_rate": 2.983793562696423e-05, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.17586535215377808, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0381, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.1623464673757553, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.037, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.16550037264823914, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0373, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.20314498245716095, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0366, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.19387134909629822, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0377, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.17531387507915497, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0381, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.21466659009456635, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.037, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1693795919418335, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0339, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.21929262578487396, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0332, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.2253660261631012, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0339, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.2084304839372635, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0363, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.2513981759548187, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0357, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.19585898518562317, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0353, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.22262930870056152, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0357, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.19813011586666107, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0347, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.16626323759555817, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0346, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.271424800157547, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.036, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2283792346715927, + "learning_rate": 2.982684510938631e-05, + "loss": 0.04, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.15996390581130981, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0394, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.20269536972045898, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0387, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.20575514435768127, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0386, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.15178649127483368, + "learning_rate": 2.982433099619666e-05, + "loss": 0.037, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.23187878727912903, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0355, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.12988153100013733, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0369, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19367633759975433, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0367, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.15319304168224335, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0337, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.19311629235744476, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0383, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.14771778881549835, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0393, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.20381954312324524, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0414, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22292324900627136, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0358, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.13758333027362823, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0359, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.18837112188339233, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0339, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.1757460981607437, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0352, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.24256370961666107, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0346, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.2417164295911789, + "learning_rate": 2.981603574229741e-05, + "loss": 0.037, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.21129392087459564, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0371, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16878797113895416, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0367, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.17526096105575562, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0353, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.1812993884086609, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.037, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.16877815127372742, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0353, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.20823006331920624, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0384, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1536571979522705, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0356, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.1477498710155487, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0346, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.12916848063468933, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0318, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.15418170392513275, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0369, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.1883322149515152, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0369, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.1569332480430603, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0374, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.21140694618225098, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0365, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.26859667897224426, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0395, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.2182101309299469, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0382, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.18896763026714325, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0372, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.17398814857006073, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0338, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.14043423533439636, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0361, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.14995983242988586, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0363, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.15542328357696533, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0361, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.16333894431591034, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0337, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.18489807844161987, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0352, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.1794673502445221, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.037, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.1882123351097107, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0414, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.16483181715011597, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.04, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.1727719008922577, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0361, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.1968923956155777, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0389, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1838955134153366, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.034, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.23556186258792877, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0369, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.2518751323223114, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0331, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18682284653186798, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0371, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.2049458622932434, + "learning_rate": 2.979480675356209e-05, + "loss": 0.04, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.11846950650215149, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0347, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.21428292989730835, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0354, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.154558002948761, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0359, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1178995817899704, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0364, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.16926616430282593, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0373, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.1721247136592865, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0346, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.1975203901529312, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0351, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.17340779304504395, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0347, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.22424420714378357, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0384, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.1688208281993866, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.16542011499404907, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0343, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.15097306668758392, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0372, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.15530438721179962, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0345, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.16840992867946625, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.034, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.19959625601768494, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0375, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.15717248618602753, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0344, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.21473488211631775, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.04, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.18634898960590363, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0346, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.16032545268535614, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0377, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.19153323769569397, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0408, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.16018801927566528, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0334, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.1498652696609497, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0349, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.1652575582265854, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0344, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.15636402368545532, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0331, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.22963644564151764, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0359, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.16931135952472687, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0389, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.17646029591560364, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0322, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.19198739528656006, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0318, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.16826699674129486, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0389, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.1693607121706009, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.037, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.147409588098526, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0343, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.1984914094209671, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0329, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.18371893465518951, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0357, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.16279958188533783, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0377, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2122160941362381, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0367, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.20099543035030365, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0343, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.17249372601509094, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0325, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1106863021850586, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0336, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.2338913530111313, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0331, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.1545400768518448, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0321, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.17114387452602386, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0323, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.16218030452728271, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.037, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.14618389308452606, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0306, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.17804935574531555, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0348, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1512002944946289, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0318, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.14877542853355408, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.037, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.1693819761276245, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0325, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.17713536322116852, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0356, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.1719064563512802, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0358, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.19448526203632355, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0363, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.18047156929969788, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0326, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.20896470546722412, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0362, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.15999653935432434, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0318, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.17288383841514587, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0336, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.21483922004699707, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0367, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.19133612513542175, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0336, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.16010481119155884, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0377, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.1778886914253235, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0354, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.14997199177742004, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0326, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.16331784427165985, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0326, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.18247577548027039, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0367, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1838751882314682, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0338, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.17194764316082, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0365, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.1354181170463562, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0315, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.19473528861999512, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0345, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14786818623542786, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0354, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.1931883543729782, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0345, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.159915491938591, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0357, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14373569190502167, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0309, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.1468040943145752, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0339, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.1592605859041214, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0374, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.17073556780815125, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0365, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.13732466101646423, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0331, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.18718381226062775, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0337, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.179783895611763, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0383, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.15795794129371643, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0368, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.13824236392974854, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0343, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.2081076204776764, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0385, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.1407512128353119, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0338, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.1619846224784851, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0325, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.184702068567276, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0353, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.18360818922519684, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0349, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.17683662474155426, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0332, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.17690211534500122, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0364, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.18250274658203125, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0336, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.12902002036571503, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0344, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.1598006635904312, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0332, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.17691342532634735, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.037, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.2213180959224701, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0324, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.14762577414512634, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0321, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.14543993771076202, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.034, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.18970227241516113, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0377, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.15049827098846436, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0341, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.1485714316368103, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0361, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1893187314271927, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0354, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.2345002293586731, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0336, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.24119973182678223, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0325, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.18577414751052856, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0337, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.1941320151090622, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0355, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.17646151781082153, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0342, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.13801687955856323, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0351, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.15124225616455078, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0337, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19114945828914642, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.033, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.23505449295043945, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0357, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.15116260945796967, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.15656375885009766, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0322, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.18982963263988495, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0368, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.18300476670265198, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0379, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.16485247015953064, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0335, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.19399121403694153, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0353, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.22944431006908417, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0343, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.16755999624729156, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0351, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.15779320895671844, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0318, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.2113935351371765, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0343, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.14698673784732819, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0332, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19958600401878357, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0357, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20250289142131805, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0334, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.1955910474061966, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0381, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.13143742084503174, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0321, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.15840105712413788, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0333, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.20588421821594238, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0347, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.1881185919046402, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0345, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.16714425384998322, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.032, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.16526015102863312, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0316, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.1865321844816208, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0351, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.19202300906181335, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0325, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.15551044046878815, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0339, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.15684190392494202, + "learning_rate": 2.969840338753429e-05, + "loss": 0.034, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.16806578636169434, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0319, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.1554284542798996, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0348, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1836974322795868, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0375, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.20003411173820496, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0364, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.1747797578573227, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0328, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.2204974740743637, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0331, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.17183512449264526, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0381, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.14398148655891418, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.03, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.17428989708423615, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0329, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.13864420354366302, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0319, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.15052364766597748, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0357, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.21513856947422028, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0349, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19239531457424164, + "learning_rate": 2.968758727655112e-05, + "loss": 0.0377, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.22284342348575592, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0358, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.17206035554409027, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0349, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.1991468220949173, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0316, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.13542494177818298, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0329, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.14473386108875275, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0346, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.1972355991601944, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0348, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.17447993159294128, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0337, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.14419333636760712, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0355, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.19937965273857117, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0381, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.18988190591335297, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0331, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.1552465558052063, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0333, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.1769426017999649, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0344, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.17652767896652222, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0354, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1598699688911438, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0354, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.17556554079055786, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0372, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.18424467742443085, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0372, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.1591036170721054, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.038, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.1973133534193039, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0332, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.139867901802063, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0336, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.17536301910877228, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0339, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16841425001621246, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0319, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.24087557196617126, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.1419876366853714, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0341, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0315, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.14541317522525787, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0326, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.18411535024642944, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0382, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1769087016582489, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0355, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.16459472477436066, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0354, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.1770380288362503, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.034, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.16219645738601685, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0316, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.18393918871879578, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0366, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.1436525285243988, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0318, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.18282611668109894, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0376, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.16712592542171478, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0367, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.13408221304416656, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0305, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.17664852738380432, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0312, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.13730520009994507, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0323, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1278112679719925, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0322, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.1742715984582901, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0339, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.15022823214530945, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0348, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.17816156148910522, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0348, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.18016202747821808, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0357, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.21158602833747864, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0331, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1747511923313141, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0301, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.16840165853500366, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0363, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.1507093459367752, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0352, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.2081453651189804, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.035, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.1459163874387741, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0331, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.17468400299549103, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0339, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.15824441611766815, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0309, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.19502659142017365, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0305, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.1502636969089508, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.032, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.13870884478092194, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0356, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1596919149160385, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0347, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.1830880492925644, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0359, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.13567551970481873, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0319, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.12907959520816803, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0327, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.17357082664966583, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.0308, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.20002517104148865, + "learning_rate": 2.963613172841619e-05, + "loss": 0.035, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.13852699100971222, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0353, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.15445692837238312, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.11687827110290527, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0319, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.15454575419425964, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0366, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.1916663944721222, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0327, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.16530223190784454, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0309, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17331530153751373, + "learning_rate": 2.962976980219581e-05, + "loss": 0.032, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.17225238680839539, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0323, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.16390499472618103, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0324, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.16459178924560547, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0315, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.1416798084974289, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0348, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.1546057015657425, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0329, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15206973254680634, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0304, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.18791860342025757, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.03, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.17669418454170227, + "learning_rate": 2.962243236871436e-05, + "loss": 0.034, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.1335674524307251, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0323, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.17903365194797516, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0327, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.17771181464195251, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0341, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1532796174287796, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0328, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.18777944147586823, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0343, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.2553282082080841, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0342, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.24115385115146637, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0347, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.18773671984672546, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0324, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.18168678879737854, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0355, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.14390279352664948, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0323, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.13260799646377563, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0336, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.16469483077526093, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0341, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.1618199497461319, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0305, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.15524879097938538, + "learning_rate": 2.960942086834339e-05, + "loss": 0.034, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.14689622819423676, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0352, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.17949312925338745, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0329, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.2376648485660553, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0326, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.1302678883075714, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0356, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.16316285729408264, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0318, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.14314667880535126, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0347, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.13162434101104736, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0335, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.13411720097064972, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0334, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.20853494107723236, + "learning_rate": 2.960094148810064e-05, + "loss": 0.036, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.15634609758853912, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0314, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.16109681129455566, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0323, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.15969422459602356, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0292, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.13126519322395325, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0338, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.13577662408351898, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0323, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13226218521595, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0307, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.12283602356910706, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0305, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16819486021995544, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0316, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.1443309634923935, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.031, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.17339679598808289, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0335, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.12366446852684021, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0319, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.18073806166648865, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0355, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.16230227053165436, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.033, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.15816937386989594, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0331, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.15911494195461273, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0296, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.19500406086444855, + "learning_rate": 2.958564529914256e-05, + "loss": 0.034, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.16553588211536407, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0348, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.1780945360660553, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0297, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16839291155338287, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0317, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.26763999462127686, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0335, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.1292608082294464, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0319, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19598962366580963, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0346, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.16073159873485565, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0317, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.2037622481584549, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0324, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.11946277320384979, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0329, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.12799547612667084, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0329, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.23773977160453796, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0327, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.19241386651992798, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0371, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.15968935191631317, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0335, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.1712830811738968, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0302, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.16464142501354218, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0346, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.2122460901737213, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0319, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.16571341454982758, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.033, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1610543429851532, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0347, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14235250651836395, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0322, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.14999081194400787, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0299, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.033, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.1448807269334793, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0307, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.20123368501663208, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0315, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.16191548109054565, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.031, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.1475522518157959, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.17617955803871155, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0323, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.20633190870285034, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0287, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.13562145829200745, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0347, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.15800844132900238, + "learning_rate": 2.955719840324779e-05, + "loss": 0.034, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.1652759164571762, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0291, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.11819476634263992, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0309, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.17562547326087952, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0327, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.1934351772069931, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0342, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.1565367877483368, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.033, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.21048930287361145, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0357, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.16762706637382507, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0341, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.1267809122800827, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0322, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.18461033701896667, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0357, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.13451576232910156, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0327, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.15478897094726562, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0311, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1461619883775711, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0314, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.1432616114616394, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0339, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.14061163365840912, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0337, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.1950581669807434, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0315, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.13942912220954895, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.031, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.15901322662830353, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0324, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.19862298667430878, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0322, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.15155215561389923, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0302, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16394968330860138, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0299, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.13225089013576508, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0305, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.18203529715538025, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0319, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.15323573350906372, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0346, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.18822363018989563, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.038, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1510896533727646, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0336, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.12274178862571716, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.032, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.1287696361541748, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0341, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.16619601845741272, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0321, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.12426833063364029, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0298, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.1751766800880432, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0341, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.13578996062278748, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0306, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.19475996494293213, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0314, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.13291390240192413, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0341, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.16193236410617828, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0325, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.16108311712741852, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0356, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.16233862936496735, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0303, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.14042457938194275, + "learning_rate": 2.951955450025455e-05, + "loss": 0.031, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.14922280609607697, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0336, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.1389254927635193, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0324, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.21042059361934662, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0331, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.18529878556728363, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0311, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1345708668231964, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0316, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.14794935286045074, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0367, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.18609392642974854, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0326, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.1762595772743225, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0312, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.15549533069133759, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0342, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.1910371482372284, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0316, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.12789981067180634, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0286, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.11457879841327667, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0303, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.12534750998020172, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0313, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.17089471220970154, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0318, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.16774669289588928, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0326, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.18693485856056213, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0356, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.20419354736804962, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15851624310016632, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0328, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1284184604883194, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.1148369237780571, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0291, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.15302640199661255, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.037, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.16082480549812317, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.031, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.16006684303283691, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0346, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14866025745868683, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0323, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.15490789711475372, + "learning_rate": 2.949326462724779e-05, + "loss": 0.033, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.18960629403591156, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0302, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.16329076886177063, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0338, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1700190156698227, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0326, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.14541998505592346, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0332, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.13809162378311157, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.16794940829277039, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0345, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1897762268781662, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0304, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2064688503742218, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0332, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.14193041622638702, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0332, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.15571711957454681, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0339, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.14264710247516632, + "learning_rate": 2.948147907679037e-05, + "loss": 0.036, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.1411963850259781, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0344, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.18891720473766327, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0308, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.15993206202983856, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0339, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.15609434247016907, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0315, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.19272638857364655, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0325, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.2836095690727234, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0323, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16382543742656708, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0335, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.18899215757846832, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0301, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.17764721810817719, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0322, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.1808568239212036, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0304, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.166172593832016, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0298, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.18262767791748047, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0305, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.17854458093643188, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.033, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.16440987586975098, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.033, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.19478020071983337, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0288, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.2605016231536865, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0318, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1850014477968216, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0313, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.1339086890220642, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.17828164994716644, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0315, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.1235261932015419, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0328, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.13704872131347656, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0322, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1296597719192505, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0315, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.16906750202178955, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0347, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14643864333629608, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.031, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.16102853417396545, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0315, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.17575663328170776, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0312, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1488381028175354, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0318, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.14629733562469482, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0333, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.17051763832569122, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.034, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.14636048674583435, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0315, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.19957765936851501, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.032, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.15027561783790588, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0342, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1580146700143814, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.032, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1452188491821289, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0302, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.1924576759338379, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0323, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.199595108628273, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0326, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.1557701975107193, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0303, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.21232250332832336, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0324, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.13840343058109283, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0321, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.14219941198825836, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0314, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1747855842113495, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0342, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1730964183807373, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0334, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.12938249111175537, + "learning_rate": 2.94341324778745e-05, + "loss": 0.029, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1782408058643341, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0299, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.943188093053675e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.12131186574697495, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0277, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.19141671061515808, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0307, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.2431664764881134, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0326, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15145368874073029, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0329, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.19781085848808289, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0332, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.19477561116218567, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0317, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.1818031370639801, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0352, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.13725785911083221, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0272, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.1874770075082779, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0322, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.18701452016830444, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0304, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.15173715353012085, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.032, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.1315767467021942, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0331, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.16627946496009827, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0299, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.2122701108455658, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0308, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.12093837559223175, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0303, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.2072226107120514, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0331, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.14395366609096527, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0323, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.18454021215438843, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0349, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1647966057062149, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0342, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1907777190208435, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0347, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.14438189566135406, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0295, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.16802345216274261, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0283, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.1488419622182846, + "learning_rate": 2.940567320758981e-05, + "loss": 0.033, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14881044626235962, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0318, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.18762606382369995, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0322, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.15426765382289886, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0314, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.17882433533668518, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0361, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.16833186149597168, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0319, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.16899487376213074, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0283, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.15301142632961273, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0325, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.13270439207553864, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0305, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.16603541374206543, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0347, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.21959784626960754, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0299, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.24568438529968262, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0339, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.14889934659004211, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0284, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.20354624092578888, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0309, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.18124797940254211, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0352, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.13871392607688904, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0311, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.18997471034526825, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0299, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.09903991222381592, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0304, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.13336266577243805, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0306, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.1564595103263855, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0295, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.17495550215244293, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.17048844695091248, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0266, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.1257491409778595, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0299, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.15244479477405548, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0341, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.15737923979759216, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.03, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.13226579129695892, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0268, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.21549297869205475, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0323, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.2102872133255005, + "learning_rate": 2.937416924082395e-05, + "loss": 0.029, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.14369380474090576, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0306, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.1390914022922516, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0273, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.13536471128463745, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0299, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.13321688771247864, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0298, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13358351588249207, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0309, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.1732599437236786, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0319, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.17455634474754333, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0314, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.1714484691619873, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0345, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.18120428919792175, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0293, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.21681863069534302, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0335, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.12099924683570862, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0299, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.13233523070812225, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0372, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.16708864271640778, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0305, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.2122814804315567, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0319, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14159013330936432, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0343, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.13396364450454712, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0308, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.17630934715270996, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0349, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.16417403519153595, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0316, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.16593796014785767, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0291, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.13613900542259216, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0299, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.15980054438114166, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0319, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1578991562128067, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0288, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.1628173291683197, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0317, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.15937024354934692, + "learning_rate": 2.934549758921277e-05, + "loss": 0.031, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.15258124470710754, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0323, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.16142596304416656, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0289, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.15643659234046936, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0295, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.17128989100456238, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0313, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.1871197372674942, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0314, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.1448175013065338, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0318, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.21338209509849548, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0334, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15262271463871002, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0293, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.14302000403404236, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0307, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.13756762444972992, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0288, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.1390128880739212, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0286, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.16088609397411346, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0299, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.11852050572633743, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0284, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.17093200981616974, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0331, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.2325763702392578, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0269, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.19948336482048035, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0314, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.16918544471263885, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0311, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.14563684165477753, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0301, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.19645430147647858, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0312, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.18950732052326202, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0292, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.11659497767686844, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0307, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1314595341682434, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0326, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1457567662000656, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.16525784134864807, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0322, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.15977199375629425, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0334, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.17081087827682495, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0308, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.20844970643520355, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.029, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.22166086733341217, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1766587346792221, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0298, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1307239532470703, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0298, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0289, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.17264464497566223, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0317, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.15076857805252075, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0308, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14443998038768768, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0291, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.13115015625953674, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0329, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1444655954837799, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0294, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.16740760207176208, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0318, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.14201150834560394, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0314, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.14225733280181885, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0297, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.16758595407009125, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0307, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.16751281917095184, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.03, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.1855877935886383, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0333, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.17420977354049683, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0277, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.1568649560213089, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0312, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.15874658524990082, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0297, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.13727585971355438, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0334, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.14999760687351227, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0328, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.16789689660072327, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.14890174567699432, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0312, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.18001733720302582, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0322, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.1718987226486206, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.03, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.18727312982082367, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0297, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.1528918594121933, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0296, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.13372628390789032, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0288, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.14980314671993256, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0319, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.11368083208799362, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0312, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.14968374371528625, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0311, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.15593667328357697, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0287, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.1762111783027649, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0296, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.15493011474609375, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0272, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.14142747223377228, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.03, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.1307222545146942, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0294, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.1598598062992096, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.03, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.17966923117637634, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0337, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15533174574375153, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0276, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.1223452091217041, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0294, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.12733329832553864, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0289, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.15903618931770325, + "learning_rate": 2.926085774098786e-05, + "loss": 0.029, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1539735347032547, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0272, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.20487286150455475, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0312, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.19920656085014343, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0315, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.15179695188999176, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0305, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.16460831463336945, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0315, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.21031928062438965, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0283, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.13835182785987854, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0287, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.1597253829240799, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0336, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.11488606780767441, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0328, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.18707934021949768, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0328, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.18874959647655487, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0317, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.1182878389954567, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0316, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.11695742607116699, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0308, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.1597018986940384, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0324, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.17729975283145905, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0316, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.18250048160552979, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0309, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.20568659901618958, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0293, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.12907297909259796, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0281, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.1809365600347519, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0333, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.17379151284694672, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0327, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.16082972288131714, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0308, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.17415443062782288, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0299, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.1659587323665619, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0272, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.1538572907447815, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0309, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.16303570568561554, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0322, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1444631963968277, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0306, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.24458004534244537, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0285, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.15384158492088318, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0277, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1425933837890625, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0303, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.12485234439373016, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0279, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.14287105202674866, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0297, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.1621903032064438, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0268, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.145540252327919, + "learning_rate": 2.921797581830774e-05, + "loss": 0.03, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.1410767287015915, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0309, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.10953503102064133, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0299, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.15029482543468475, + "learning_rate": 2.92140191009304e-05, + "loss": 0.031, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.13262124359607697, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0299, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1503380835056305, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0329, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.18932904303073883, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.028, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.1473148614168167, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0315, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.18515919148921967, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0309, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.16122235357761383, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0279, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.1684419959783554, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0309, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.21400189399719238, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0304, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.12552157044410706, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0286, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.15383073687553406, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.029, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.1269458532333374, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0289, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.1839684545993805, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0316, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.13441112637519836, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0281, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.13045959174633026, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0296, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.13933542370796204, + "learning_rate": 2.919408984114327e-05, + "loss": 0.029, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.15143601596355438, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0296, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.16731251776218414, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.032, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.13245129585266113, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0268, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.17244738340377808, + "learning_rate": 2.918873439576597e-05, + "loss": 0.028, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.1709858626127243, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0293, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18727441132068634, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0302, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1482084095478058, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0297, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.1343366652727127, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0303, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.1456393450498581, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0299, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1569104641675949, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0281, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.19711613655090332, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0298, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.14027468860149384, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.031, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.16852174699306488, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0289, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.18446287512779236, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0308, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.16207733750343323, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0283, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.19747264683246613, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0301, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.15206509828567505, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0295, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.18873688578605652, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0286, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.144051656126976, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0335, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.14620175957679749, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0284, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.19520114362239838, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0304, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.16127106547355652, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0333, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.11467311531305313, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0303, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11777719110250473, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0311, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.12151405215263367, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0278, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1852089911699295, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0297, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.19555963575839996, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0321, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.1584533303976059, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0292, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.1764054298400879, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0306, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.16257385909557343, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0298, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.13896049559116364, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0296, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.16509538888931274, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0321, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.14912301301956177, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0285, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.18058481812477112, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0302, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.18734610080718994, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0289, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.19707795977592468, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0327, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.16215947270393372, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0302, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1710345447063446, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0332, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.19638675451278687, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0316, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.17807847261428833, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0285, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.13119569420814514, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0298, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.13414712250232697, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0301, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.1941734254360199, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0327, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.19624412059783936, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0291, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.17615875601768494, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0293, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.19011878967285156, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0316, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.15080299973487854, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0263, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.1676572859287262, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0318, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.14975300431251526, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0263, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17691564559936523, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.17775340378284454, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0298, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.14438258111476898, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.14729741215705872, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0289, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.20924119651317596, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0297, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.15471352636814117, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0289, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.1571032702922821, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0307, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15048687160015106, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0301, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.15063059329986572, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0279, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15528838336467743, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0299, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.1756942719221115, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0291, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.1491880565881729, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0296, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.1562451720237732, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0302, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.12366753816604614, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0257, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.15308785438537598, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.16614912450313568, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0301, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.12428635358810425, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0303, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.135500967502594, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0283, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.16944879293441772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0279, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.174363911151886, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0295, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.16406473517417908, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0284, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.2144644856452942, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0277, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.22006124258041382, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0264, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.15261462330818176, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0298, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.15205760300159454, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.14101643860340118, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0305, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16015839576721191, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.17772945761680603, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0312, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16818812489509583, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0283, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1448637843132019, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0325, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.14702047407627106, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0303, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.15547925233840942, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0299, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.18527188897132874, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.029, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.1533391922712326, + "learning_rate": 2.907944011202334e-05, + "loss": 0.029, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.15355326235294342, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0286, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.1503538340330124, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0312, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.12572433054447174, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0277, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.13614071905612946, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0303, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.13363786041736603, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0292, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.13318102061748505, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.0289, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.14687903225421906, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0301, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.13144652545452118, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0271, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.16351011395454407, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0292, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.19256997108459473, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0268, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13006120920181274, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0285, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.12932874262332916, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.11981740593910217, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.14302268624305725, + "learning_rate": 2.905937392537971e-05, + "loss": 0.029, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.1446913331747055, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0289, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.1500432789325714, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.15604284405708313, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0283, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.1795264482498169, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0289, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.14356553554534912, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0304, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.1323058009147644, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0302, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.2098175585269928, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0297, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.1424631029367447, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0295, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.1956447958946228, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0294, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.11993541568517685, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0273, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.18414588272571564, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0316, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.17320625483989716, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0319, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.17650462687015533, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0304, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.18740026652812958, + "learning_rate": 2.903909846652783e-05, + "loss": 0.034, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.12898528575897217, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0309, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.11721221357584, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0297, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.1659565567970276, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0294, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.1469241976737976, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0298, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14588896930217743, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0332, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.16243143379688263, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0305, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.15857404470443726, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0318, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.14992372691631317, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0305, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.13760966062545776, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0299, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.14302772283554077, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0313, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.18560320138931274, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0262, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.13117150962352753, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0286, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.16370882093906403, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0298, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.18689852952957153, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.031, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.16412676870822906, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0289, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.20153313875198364, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0297, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.15255704522132874, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0316, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.2036312222480774, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0306, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.16087041795253754, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0315, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.1514595001935959, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0295, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.13961558043956757, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0259, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1996241807937622, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0336, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.15389376878738403, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0278, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.12345903366804123, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.032, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.16360828280448914, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0292, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.19773486256599426, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.032, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.11931311339139938, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0268, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.1378420740365982, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0296, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.09076640754938126, + "learning_rate": 2.899643488777395e-05, + "loss": 0.026, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.1116260215640068, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0266, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.18670536577701569, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0314, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.15351636707782745, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0279, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.15310993790626526, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0298, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.15458840131759644, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0329, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.12034284323453903, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0266, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.14272604882717133, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0296, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.16335424780845642, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0266, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.16105882823467255, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0355, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.1455371230840683, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0261, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.21555180847644806, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0301, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.15278705954551697, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0292, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.14099596440792084, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0287, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.10848120599985123, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.028, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.1380046308040619, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0273, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.14006251096725464, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0328, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.16518674790859222, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0286, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.13295377790927887, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.031, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.1623251736164093, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0281, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.16029980778694153, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.03, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.15575948357582092, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0294, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.157670795917511, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0294, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1470523178577423, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0281, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.12192463129758835, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0248, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.1240374892950058, + "learning_rate": 2.895893835994484e-05, + "loss": 0.028, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1385754942893982, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0287, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.140727236866951, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0281, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.15758906304836273, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0294, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.19664499163627625, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0277, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.12567110359668732, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0279, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.13595536351203918, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0313, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.14262230694293976, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0317, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1474338322877884, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0282, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.18601401150226593, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0274, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.1392786204814911, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0269, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.12022347748279572, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0262, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.17340584099292755, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0277, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.1438349187374115, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0273, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.1408676654100418, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0277, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.14853249490261078, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0291, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.20401965081691742, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.031, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.18385861814022064, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.12695786356925964, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0267, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.14115597307682037, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0291, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.15790492296218872, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0294, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.15104074776172638, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0278, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13686735928058624, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0276, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.12029985338449478, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0305, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1351509988307953, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0293, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.10609307885169983, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0285, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15056312084197998, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0292, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.17062142491340637, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0281, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.15565407276153564, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0298, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.136429563164711, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0282, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.1649797409772873, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0267, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.19815988838672638, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.032, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15660062432289124, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.03, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.1840171217918396, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0342, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.14063017070293427, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0282, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.14935986697673798, + "learning_rate": 2.890533053632388e-05, + "loss": 0.029, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.14691074192523956, + "learning_rate": 2.89037798520224e-05, + "loss": 0.033, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.14992092549800873, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0262, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.16885656118392944, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0301, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1223178282380104, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0275, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.1200852021574974, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0277, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.1251501888036728, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0252, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.16905754804611206, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0277, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.13312722742557526, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0296, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.13198228180408478, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0269, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.14326612651348114, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0253, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.14861920475959778, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0262, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.19368910789489746, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0295, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.143910750746727, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0282, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.12314137071371078, + "learning_rate": 2.888352491214935e-05, + "loss": 0.027, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.16095006465911865, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0272, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.1414606273174286, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.029, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.15631091594696045, + "learning_rate": 2.887882538107974e-05, + "loss": 0.0291, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16476406157016754, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0283, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.2236901968717575, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0286, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.1318763792514801, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0274, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.15661245584487915, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.029, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.13406990468502045, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0287, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16861270368099213, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0272, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.18082283437252045, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0285, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.17517460882663727, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0277, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.17011325061321259, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0325, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14378181099891663, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0306, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1766553819179535, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0304, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.1666528880596161, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0294, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.15178363025188446, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0273, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.14275510609149933, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0291, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.1568918377161026, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0267, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.1370164304971695, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0268, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.10388749092817307, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0289, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11448585987091064, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0302, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.15148931741714478, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0292, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.14388935267925262, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0257, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.14465516805648804, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0272, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.13414210081100464, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0273, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12774589657783508, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.028, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.24874050915241241, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0282, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.1624344289302826, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0294, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.12817884981632233, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0324, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1509743630886078, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0269, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.31345340609550476, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0287, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.1523321270942688, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0259, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.1633133739233017, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0269, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.14278745651245117, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.0253, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.18361203372478485, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0282, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.10244551301002502, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0268, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.125647634267807, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0264, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.12238023430109024, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0264, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.16535502672195435, + "learning_rate": 2.882169202177678e-05, + "loss": 0.0329, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.19624854624271393, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0303, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.18285983800888062, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0266, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.13460052013397217, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.0293, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.1427820920944214, + "learning_rate": 2.881525979398615e-05, + "loss": 0.0244, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.14855900406837463, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0276, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.1433439403772354, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0299, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.1452850103378296, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0271, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.14403793215751648, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0264, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.13841332495212555, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0277, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.1426801085472107, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0296, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.14492706954479218, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0272, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.1315297782421112, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.0289, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.1513436734676361, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0284, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.15784604847431183, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.0273, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.12729156017303467, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0297, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.20208381116390228, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0284, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.1486770659685135, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0271, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.1463368982076645, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.0292, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.12362723797559738, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0252, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.18984590470790863, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0292, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.12962038815021515, + "learning_rate": 2.878773564250086e-05, + "loss": 0.0301, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.132072314620018, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0298, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.1482372134923935, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0267, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.15687556564807892, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0308, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.11656040698289871, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.028, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.16482318937778473, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0256, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.13332504034042358, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.027, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.16453887522220612, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0253, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.16902299225330353, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0283, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.12483953684568405, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0276, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.14033488929271698, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0277, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.18922871351242065, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0258, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.15026360750198364, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0298, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17492659389972687, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0275, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.14793051779270172, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0301, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.17575892806053162, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.027, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.1661873608827591, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0275, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.14529843628406525, + "learning_rate": 2.875990888327069e-05, + "loss": 0.0277, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.11084084957838058, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0278, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1433780938386917, + "learning_rate": 2.875661528053842e-05, + "loss": 0.026, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.13195206224918365, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0298, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.14103750884532928, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0309, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.17067542672157288, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0296, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.14078323543071747, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0288, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.17906904220581055, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0268, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.163010373711586, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0285, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.17636623978614807, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0276, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.18726587295532227, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0297, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.18583180010318756, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.16773444414138794, + "learning_rate": 2.874008460348417e-05, + "loss": 0.0289, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.14579996466636658, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.12655268609523773, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0246, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.15124735236167908, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0297, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.16273140907287598, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0295, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14014655351638794, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0262, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12657561898231506, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0292, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.13813915848731995, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.0287, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.12874144315719604, + "learning_rate": 2.872678492608539e-05, + "loss": 0.026, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.12086840718984604, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0245, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.15290603041648865, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0297, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.13261260092258453, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0263, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.17371073365211487, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0268, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.17300288379192352, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0264, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.16441872715950012, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0279, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.17883789539337158, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.03, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.14519335329532623, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0266, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.18063998222351074, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0275, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.19818824529647827, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0256, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1695982664823532, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0274, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.16905854642391205, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0277, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.1852719783782959, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.03, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.14601513743400574, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0261, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.11720956116914749, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0279, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.14177089929580688, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0268, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.14933669567108154, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0292, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15183576941490173, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0285, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.1255270391702652, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.025, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.1467735320329666, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0258, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.13636574149131775, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0278, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.12642863392829895, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0254, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.09835807979106903, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0284, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.11784205585718155, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0285, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.14453290402889252, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0301, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.19569514691829681, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0264, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.13860125839710236, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.026, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.16651827096939087, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0283, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.13682647049427032, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0309, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.14632774889469147, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0275, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.1401532143354416, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0261, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.12643243372440338, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0268, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1152961477637291, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0259, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.14027607440948486, + "learning_rate": 2.866951776928e-05, + "loss": 0.0306, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.12379784137010574, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0272, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.15595580637454987, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0288, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.12823393940925598, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0263, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1481468379497528, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0283, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.12768998742103577, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0262, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.15771906077861786, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0267, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.18978171050548553, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0295, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.1386374831199646, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0277, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.1583528220653534, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0273, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.09999477118253708, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0253, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.14432361721992493, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0289, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.19001086056232452, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.0295, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.13931159675121307, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0257, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.1621151715517044, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0269, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.15239235758781433, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0275, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12619581818580627, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0273, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.14698320627212524, + "learning_rate": 2.864043385528954e-05, + "loss": 0.0275, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.14121954143047333, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0271, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.16203583776950836, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0265, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.15562862157821655, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0267, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.16438758373260498, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0285, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.13769859075546265, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0274, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.14908984303474426, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0256, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.14057780802249908, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.0261, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.16056638956069946, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.0267, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.12299710512161255, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0257, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.12815117835998535, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0284, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.14666494727134705, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0303, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.15829558670520782, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0243, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1499258577823639, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.029, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.16584458947181702, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0272, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1765419989824295, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0278, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.17066679894924164, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0281, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13183729350566864, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0255, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.12179457396268845, + "learning_rate": 2.86093128283797e-05, + "loss": 0.028, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.10134928673505783, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0275, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16393089294433594, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0278, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1360260546207428, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0297, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.16158054769039154, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0249, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.12037541717290878, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.0286, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.16677190363407135, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0277, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1442420333623886, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0268, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.1756538599729538, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0282, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.11580783128738403, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0282, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.11671322584152222, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0279, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.15163642168045044, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0297, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.1390339583158493, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0282, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.11784064769744873, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0275, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.15270282328128815, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0282, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.09030280262231827, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0281, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.15108083188533783, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0274, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.1403200924396515, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0265, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.14890284836292267, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0303, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.11579708009958267, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0283, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.14590834081172943, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0268, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.1513289511203766, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0264, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.13494084775447845, + "learning_rate": 2.857082133984191e-05, + "loss": 0.026, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.11738834530115128, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.029, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.1497882753610611, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0271, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.17350555956363678, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0277, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.16718043386936188, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0269, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.12658995389938354, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0267, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15880689024925232, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0256, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.1480095237493515, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0263, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.13779640197753906, + "learning_rate": 2.85567006866004e-05, + "loss": 0.0282, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1088612824678421, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0245, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.17420272529125214, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0267, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.12365500628948212, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.025, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.13261641561985016, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0256, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.10485266894102097, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0253, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14070947468280792, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0284, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.1179763451218605, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0241, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13687312602996826, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0284, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.17160922288894653, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0263, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.17487400770187378, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0275, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.11352817714214325, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1871306300163269, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0277, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.13434743881225586, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0278, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.16361933946609497, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0257, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.19457586109638214, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0277, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.16905367374420166, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0323, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.14845560491085052, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0248, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.12934044003486633, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.026, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.11046823114156723, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0267, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.14174290001392365, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0255, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.15194296836853027, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.0257, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.1532754749059677, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0269, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19864243268966675, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0288, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.16765300929546356, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0271, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.15602844953536987, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0288, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.1571321040391922, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0264, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1625281423330307, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0268, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.12808986008167267, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0307, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.12471310794353485, + "learning_rate": 2.850496147627652e-05, + "loss": 0.027, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.1521507054567337, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0289, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1389119178056717, + "learning_rate": 2.850136142292186e-05, + "loss": 0.024, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1338133066892624, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13214652240276337, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.0274, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.17002162337303162, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0277, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.1433296948671341, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0267, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.19547788798809052, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0287, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.1916888952255249, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0296, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.17188303172588348, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0262, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.14335697889328003, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.0269, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14298376441001892, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0278, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.12892583012580872, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0278, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.16246432065963745, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0255, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.1237327978014946, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0252, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.12249717861413956, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0238, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.16828545928001404, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0262, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.12834671139717102, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0287, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.11583935469388962, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.029, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.10479022562503815, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0244, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.18407097458839417, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0276, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.12898766994476318, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0277, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.13034385442733765, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0241, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.13197654485702515, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0269, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.14327986538410187, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0281, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.15228267014026642, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0269, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.11858481913805008, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0261, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.16075590252876282, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0284, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.1527116447687149, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0283, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.1280883252620697, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0294, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1348433941602707, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0251, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.15039700269699097, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0272, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.16682399809360504, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.0277, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.11187390983104706, + "learning_rate": 2.844503681416359e-05, + "loss": 0.0298, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.1675758808851242, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0267, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.16022244095802307, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0301, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.16192372143268585, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0277, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.13092738389968872, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.12416383624076843, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.028, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.12514744699001312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0252, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.1705007702112198, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.0274, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.14251261949539185, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0264, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.147093266248703, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0279, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.12747326493263245, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0265, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.14222867786884308, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0276, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.12574070692062378, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0262, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.12812504172325134, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0225, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.13585571944713593, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0267, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.12168630957603455, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0236, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.12048719823360443, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0225, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.15165774524211884, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0267, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.15395723283290863, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.026, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.14715734124183655, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0259, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.11370360106229782, + "learning_rate": 2.840817735753433e-05, + "loss": 0.0284, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1264869123697281, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.028, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.13726982474327087, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0268, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1450515240430832, + "learning_rate": 2.840261328154438e-05, + "loss": 0.0251, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.1328781247138977, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.025, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.15474216639995575, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.0292, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.1338219791650772, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0279, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.16368664801120758, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.027, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.18125605583190918, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.0252, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.127846822142601, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0252, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.17743080854415894, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0275, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.16253551840782166, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0287, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.14212867617607117, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0274, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.1380675882101059, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0263, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.12773451209068298, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.0257, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.1556098759174347, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0264, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.16690294444561005, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0296, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.11710768938064575, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0252, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.17441801726818085, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0318, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.10790088772773743, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0251, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.16537238657474518, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0266, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.1414048671722412, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0276, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.1096765398979187, + "learning_rate": 2.836716156259365e-05, + "loss": 0.0258, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.14450953900814056, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0278, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15733887255191803, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0278, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.15253902971744537, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0275, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.1410975307226181, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0249, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1242721900343895, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0251, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.15358713269233704, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0287, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.15856830775737762, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.029, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.1777517944574356, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0254, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.10823221504688263, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0262, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.144780695438385, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0286, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12900282442569733, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0243, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.15370796620845795, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0238, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.15801627933979034, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0259, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.19058606028556824, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0257, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14663439989089966, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0281, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.10609344393014908, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.0261, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.11403398960828781, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0291, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.11631747335195541, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0243, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.1325336992740631, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0266, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.11245301365852356, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.10258981585502625, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0252, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.15847361087799072, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0258, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.15267148613929749, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0246, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.15297843515872955, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0248, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.1767774075269699, + "learning_rate": 2.831995639121568e-05, + "loss": 0.0292, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.12696890532970428, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.025, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.17774799466133118, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0267, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.15603452920913696, + "learning_rate": 2.83142492407012e-05, + "loss": 0.0285, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.12731599807739258, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0272, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.13816066086292267, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0265, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.14216956496238708, + "learning_rate": 2.830853299002291e-05, + "loss": 0.026, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1427476406097412, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0263, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.14667609333992004, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0282, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.11785806715488434, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0276, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.125101238489151, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0255, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.10735481232404709, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0248, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.14256443083286285, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0286, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12700040638446808, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0275, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.11069332808256149, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0283, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.1438131332397461, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.0269, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.1901932805776596, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0274, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.17823027074337006, + "learning_rate": 2.828557706390495e-05, + "loss": 0.029, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.13579529523849487, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0267, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.1431250274181366, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0279, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.11313948035240173, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0274, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1669740527868271, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0288, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.12642791867256165, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0267, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.15632158517837524, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0256, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.1404937356710434, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0263, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.13466882705688477, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0249, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.13014933466911316, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0255, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.18688003718852997, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0267, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.14594615995883942, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0253, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.1708635836839676, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0244, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.16193392872810364, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0265, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.13249778747558594, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0286, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.13875705003738403, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0262, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.19137892127037048, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0246, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.15596376359462738, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0248, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.1577422320842743, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0247, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.14322742819786072, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0263, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.13651713728904724, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0261, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11288725584745407, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0252, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.09636299312114716, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0267, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.11729153245687485, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0275, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.14827747642993927, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0265, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.1446153074502945, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0257, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.1329014152288437, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0247, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.12726181745529175, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0259, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.13508957624435425, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0258, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.14538758993148804, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0264, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.10585593432188034, + "learning_rate": 2.822755216255228e-05, + "loss": 0.0276, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.17435809969902039, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0281, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.1504661589860916, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0259, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.1678706258535385, + "learning_rate": 2.8221699881734e-05, + "loss": 0.029, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.13508610427379608, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0261, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12610375881195068, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0237, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.12272591143846512, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0231, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.1211056336760521, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0252, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.13204888999462128, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0276, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.11982017755508423, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0245, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.17140068113803864, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0236, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.12660211324691772, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0264, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1636001467704773, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0266, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.16939589381217957, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0276, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.15193505585193634, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0291, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16562964022159576, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0283, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.13603557646274567, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.024, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.14359496533870697, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0257, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.14077764749526978, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0247, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.14604462683200836, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0277, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.1240646094083786, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0262, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13144497573375702, + "learning_rate": 2.818639656220465e-05, + "loss": 0.0255, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.1384861171245575, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0227, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.16475403308868408, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0269, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.14310070872306824, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0258, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1498526632785797, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0263, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.13084106147289276, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0268, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.12004188448190689, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0252, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.14093606173992157, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0262, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.11016195267438889, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.0249, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.11101678013801575, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0246, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.1335095763206482, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0274, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1479913741350174, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0237, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.11868918687105179, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0231, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12799321115016937, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.027, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.17011503875255585, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0271, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.15915420651435852, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0274, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.11501206457614899, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0254, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.1457805633544922, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0274, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.1501278430223465, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0264, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.13190113008022308, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0272, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.14392758905887604, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0235, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.1375834047794342, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0241, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.14287824928760529, + "learning_rate": 2.814280752477756e-05, + "loss": 0.0269, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12234462797641754, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0259, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.1336476355791092, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0242, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.11315035820007324, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0245, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.14132073521614075, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0293, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.12053677439689636, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.0266, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.10896297544240952, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0268, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.11100424081087112, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.026, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.11256317049264908, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0237, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.14322060346603394, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0255, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.1655588448047638, + "learning_rate": 2.812283453196085e-05, + "loss": 0.026, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.14714404940605164, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0261, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.12317050993442535, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0247, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12872155010700226, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0258, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.1666097790002823, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0261, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.13425199687480927, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.16131597757339478, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0295, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.13762447237968445, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0242, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.12883082032203674, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0274, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.15029405057430267, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0269, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.13236507773399353, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0303, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.12054237723350525, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0218, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.1471109539270401, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0242, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.11926044523715973, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0275, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.13649292290210724, + "learning_rate": 2.809470494867301e-05, + "loss": 0.0231, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.15434928238391876, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0283, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.11536315083503723, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0271, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.12312797456979752, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0249, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.12750110030174255, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0235, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.1568569540977478, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0277, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.11667264997959137, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0258, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.14598897099494934, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0266, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.21723008155822754, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0258, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.14233075082302094, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0244, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.14836373925209045, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0265, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.146131694316864, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0258, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.16575932502746582, + "learning_rate": 2.807043872632495e-05, + "loss": 0.025, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.15142853558063507, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0245, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.11762700974941254, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0253, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.09714391082525253, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0233, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.1586424708366394, + "learning_rate": 2.806231820878714e-05, + "loss": 0.023, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.14317330718040466, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.0271, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.10633813589811325, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.0244, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.19832657277584076, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0249, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.14428019523620605, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0283, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.18295790255069733, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.0257, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.16204482316970825, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0285, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.14742101728916168, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0239, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1435142159461975, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0253, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.2000540792942047, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.16311310231685638, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0253, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.1238052025437355, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0249, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.16693037748336792, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0269, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.12382039427757263, + "learning_rate": 2.803581696238626e-05, + "loss": 0.025, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.13866499066352844, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0259, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.12111294269561768, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0258, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.13717366755008698, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0266, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.1119573786854744, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0264, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.1390918791294098, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.12881262600421906, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0251, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.1354016363620758, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0256, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.1285468488931656, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0315, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.17634150385856628, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0254, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.15077067911624908, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.0268, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16464342176914215, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0265, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.3138365149497986, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0255, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.15954458713531494, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0272, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15878041088581085, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0267, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.17743942141532898, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0273, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.16661205887794495, + "learning_rate": 2.800297040373295e-05, + "loss": 0.0268, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.13536596298217773, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.0241, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.13258172571659088, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0266, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.1524200588464737, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0276, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.12779775261878967, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0245, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1008654311299324, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0222, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.14413924515247345, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0297, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.14558148384094238, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0252, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.15591812133789062, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0248, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.12246204912662506, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0276, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.13106316328048706, + "learning_rate": 2.7982312890369e-05, + "loss": 0.0265, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.12663032114505768, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0234, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.1167084351181984, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15868185460567474, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0251, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13669976592063904, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0291, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.15100552141666412, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0237, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14283764362335205, + "learning_rate": 2.796987104810136e-05, + "loss": 0.024, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.13055892288684845, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0255, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.16733981668949127, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0244, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.1716628074645996, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0257, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.11838851124048233, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0288, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.14557431638240814, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0238, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.14981132745742798, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0254, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.11149631440639496, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0256, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.20500904321670532, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0255, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.12128293514251709, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0254, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.12934274971485138, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0254, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.1355874240398407, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0267, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13485173881053925, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0259, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.1481594294309616, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0251, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.15145331621170044, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0243, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.16746139526367188, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0284, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.10760390013456345, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0268, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.1460299789905548, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0264, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.15936219692230225, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0232, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.1373136043548584, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.0255, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.14508306980133057, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0243, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.153093621134758, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0262, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.15373852849006653, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0238, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.14033931493759155, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0267, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.1424216777086258, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0252, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.11820772290229797, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0253, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.13179944455623627, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0243, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.18073157966136932, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0256, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.12908786535263062, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0254, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.18455258011817932, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0263, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.14844198524951935, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0248, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.14291873574256897, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0268, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.105900838971138, + "learning_rate": 2.79029165248073e-05, + "loss": 0.027, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.14083242416381836, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0244, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.14579042792320251, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0239, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.13705924153327942, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0254, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.17981873452663422, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0225, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.13812783360481262, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0278, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12511365115642548, + "learning_rate": 2.789025068651819e-05, + "loss": 0.027, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.13009224832057953, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0245, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.145152747631073, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0281, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.1607820838689804, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0254, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.09400778263807297, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0258, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.1608203500509262, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0263, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.14968201518058777, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0247, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.1333402544260025, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0272, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.17871612310409546, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0256, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.10688412934541702, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0264, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.154439315199852, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0259, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.1270124316215515, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0279, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.15214042365550995, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0259, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.14094167947769165, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0261, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.11419129371643066, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.025, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.1447947919368744, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0258, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.12327348440885544, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0246, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.11033894866704941, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0259, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.128106951713562, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0247, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.13206088542938232, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0254, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.13664662837982178, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0274, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1349778026342392, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0262, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.10689893364906311, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0262, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.16241611540317535, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0269, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14955222606658936, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0251, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.13519948720932007, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0231, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.15076202154159546, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0272, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.15367662906646729, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0255, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.15819156169891357, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0242, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.12844862043857574, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0259, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.13392682373523712, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0245, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.15134917199611664, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.0287, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.14139734208583832, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0254, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.1562367081642151, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0246, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.1298898160457611, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0252, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13048380613327026, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0252, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.12535928189754486, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0271, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.19301816821098328, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0245, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.12262707948684692, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0262, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1394580751657486, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0272, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.21376213431358337, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0243, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.15442071855068207, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0285, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.13063938915729523, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0277, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.15023048222064972, + "learning_rate": 2.779844963973707e-05, + "loss": 0.025, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.17523249983787537, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0239, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.1312434822320938, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0234, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13352124392986298, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0249, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.12799884378910065, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0257, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.12704412639141083, + "learning_rate": 2.778765821684581e-05, + "loss": 0.025, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.11119242757558823, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0248, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.1564178764820099, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0282, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.13500694930553436, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0279, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.1542324423789978, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0258, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13888850808143616, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0281, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.14011257886886597, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.14887000620365143, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0254, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.15670792758464813, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0289, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.1288663148880005, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0242, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.13191169500350952, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.027, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.1320677250623703, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0239, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.13418279588222504, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0234, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.11019188910722733, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0232, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.13963419198989868, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0235, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.1163327619433403, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.028, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.13508521020412445, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0275, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.11098913848400116, + "learning_rate": 2.775078589916535e-05, + "loss": 0.025, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.16348809003829956, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0259, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.12986604869365692, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0246, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1302310824394226, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0259, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.11797303706407547, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.0252, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.1427793651819229, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0233, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.13411074876785278, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.0257, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.1528121531009674, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.11762527376413345, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0238, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.13639862835407257, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0257, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.1356336623430252, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0269, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.16879825294017792, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.13352762162685394, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0236, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.20271515846252441, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0296, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.11791368573904037, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0241, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14179861545562744, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.0256, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12654612958431244, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0243, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12589605152606964, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0222, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.12714968621730804, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0278, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.13410630822181702, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0248, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.12528881430625916, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0258, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.11806205660104752, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0308, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.14317870140075684, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0244, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14099767804145813, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0228, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.15186408162117004, + "learning_rate": 2.769825463131276e-05, + "loss": 0.0252, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.18532416224479675, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0262, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.1445159614086151, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0265, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14576950669288635, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0253, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1777651309967041, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0229, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.16440069675445557, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0234, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.12858232855796814, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0259, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.14329349994659424, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0254, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.1265982836484909, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0235, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13783425092697144, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0228, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.1327633559703827, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0248, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14867408573627472, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0257, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.1556752771139145, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0239, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.16407394409179688, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0261, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.14142248034477234, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0265, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.11142456531524658, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0221, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.127807155251503, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0265, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.14716863632202148, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0241, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.1455392986536026, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.024, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1665087193250656, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0253, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.16937066614627838, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0292, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.13188236951828003, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0254, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.1010027602314949, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.0236, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.15787169337272644, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.0279, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.169437438249588, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0246, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.22705122828483582, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0227, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.17403557896614075, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.027, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.1287415474653244, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0253, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.13813741505146027, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0238, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.11655540764331818, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.13267700374126434, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.025, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.10447681695222855, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0239, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.11865647882223129, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.025, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.13245069980621338, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0256, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.11377252638339996, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0266, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.11403302848339081, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0252, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.1518944650888443, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0286, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.13505733013153076, + "learning_rate": 2.761618148370765e-05, + "loss": 0.0253, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.12230793386697769, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0247, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.1600245237350464, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0267, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.13100600242614746, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.11071992665529251, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0254, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.16208025813102722, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.0287, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.1652526706457138, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.027, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14042513072490692, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.1674395203590393, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0257, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.138246089220047, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0272, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13758982717990875, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0227, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.15503156185150146, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0259, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.15050673484802246, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0279, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.13952329754829407, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.024, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.14903564751148224, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0242, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.09984514862298965, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.09952808916568756, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0233, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.14514309167861938, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0255, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.12824438512325287, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0232, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.15507018566131592, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0251, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.12279994785785675, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0225, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.13601353764533997, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0249, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.20022229850292206, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.0263, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.1407146006822586, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0226, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.1350381076335907, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0248, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.13220156729221344, + "learning_rate": 2.755998359318217e-05, + "loss": 0.0251, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.11746657639741898, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0256, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.10933379083871841, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0235, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.1263892501592636, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0244, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.14062181115150452, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.0255, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.1430615335702896, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0254, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.19069550931453705, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0254, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.16568611562252045, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0252, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.15911205112934113, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.0265, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.17054520547389984, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0265, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.15547682344913483, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0237, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.12778308987617493, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0242, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.1590787172317505, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0232, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13945519924163818, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.023, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.14507925510406494, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0247, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14541085064411163, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0245, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.17666012048721313, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0222, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.16692468523979187, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0218, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.1295796036720276, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0242, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.12527599930763245, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0273, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.13334719836711884, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0253, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.1175302267074585, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0251, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.14461147785186768, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0261, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.15658612549304962, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0236, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.15802229940891266, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0236, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.14738211035728455, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0252, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.1639637053012848, + "learning_rate": 2.750090542643057e-05, + "loss": 0.027, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1573086827993393, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0264, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.13114139437675476, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0238, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.1112227514386177, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.027, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.12209689617156982, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.13741743564605713, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0239, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14444704353809357, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0263, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.13822096586227417, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0246, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.14305464923381805, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0237, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13539855182170868, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0264, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.13791607320308685, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0234, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.205885648727417, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0234, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.1034194752573967, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0214, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.15530072152614594, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0258, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.1535874903202057, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0226, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.19778263568878174, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0238, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.175798699259758, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0255, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.1803887039422989, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0238, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.1906212419271469, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0253, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.13742655515670776, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.026, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.12772122025489807, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0261, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13805702328681946, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0259, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.14081542193889618, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0241, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.10023088753223419, + "learning_rate": 2.744810891043728e-05, + "loss": 0.0246, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.10111884027719498, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0215, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.11498598754405975, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0217, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10912947356700897, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.0257, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.13553178310394287, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0266, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.18210133910179138, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0234, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.11901862919330597, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0248, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.14622537791728973, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0234, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.1641179323196411, + "learning_rate": 2.742962757296782e-05, + "loss": 0.026, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.14369463920593262, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.0236, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.14231407642364502, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0213, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.10929716378450394, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0216, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.13811182975769043, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0241, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.13219693303108215, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0219, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.13322590291500092, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0259, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.14716853201389313, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0255, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.11750922352075577, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0272, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.1430707424879074, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1722290813922882, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.025, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.15502606332302094, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.0231, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.09124620258808136, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0246, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.16296057403087616, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0277, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.10881830751895905, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0222, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.11321897804737091, + "learning_rate": 2.739481230505444e-05, + "loss": 0.026, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.1362120658159256, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.03, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.14329490065574646, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0252, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.4512016475200653, + "learning_rate": 2.738782381719405e-05, + "loss": 0.0238, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12559202313423157, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0263, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.12616541981697083, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0248, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.1516106277704239, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0234, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.12785211205482483, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0264, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16447393596172333, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.026, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.1648014783859253, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0264, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.14414788782596588, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0249, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.10982059687376022, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0254, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.12677313387393951, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0263, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.12355965375900269, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0223, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.15190085768699646, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0246, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.13910633325576782, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0231, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.11962535977363586, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0259, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.1566285789012909, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0256, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.12282316386699677, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0253, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.16674686968326569, + "learning_rate": 2.735040900116229e-05, + "loss": 0.025, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.14292193949222565, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0277, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.1504344791173935, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0268, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.16317704319953918, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0229, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.20625433325767517, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0242, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.1895458698272705, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.024, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.1987384408712387, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.0258, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.18668580055236816, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.1807073950767517, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0246, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.11927719414234161, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.028, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.11955022811889648, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0259, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.15261328220367432, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0268, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.15620580315589905, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0254, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.12000948935747147, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0234, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.1188279539346695, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.0248, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.13279424607753754, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0241, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.13379216194152832, + "learning_rate": 2.731275407491117e-05, + "loss": 0.026, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.12054166197776794, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0215, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.15941590070724487, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0279, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.18678002059459686, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0263, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.21807490289211273, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.0239, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.12243734300136566, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.028, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.14628900587558746, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0248, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.17471152544021606, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0257, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.11189062893390656, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0242, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.16700562834739685, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0252, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.1213584691286087, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0251, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.14496786892414093, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0258, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.17899028956890106, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0248, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13594108819961548, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0246, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.1365566849708557, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.021, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.10815171152353287, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0227, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.11390877515077591, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0209, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.1541927456855774, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0213, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.1342182159423828, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0251, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.1300208568572998, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0249, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14917100965976715, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0274, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.14349427819252014, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.0269, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10514543205499649, + "learning_rate": 2.72605878418951e-05, + "loss": 0.024, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.12946246564388275, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0234, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.1120578870177269, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.0246, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.14252731204032898, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.0287, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.0984109416604042, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0234, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.13058412075042725, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0235, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.1468561738729477, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0263, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.1297929584980011, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0224, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.15071524679660797, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0254, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.15211531519889832, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0239, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.13758648931980133, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0254, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.13972750306129456, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0239, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1762881577014923, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0235, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.13277511298656464, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.025, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.11352386325597763, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0225, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.1340995877981186, + "learning_rate": 2.722476145603374e-05, + "loss": 0.0232, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.16372360289096832, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0224, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.130090594291687, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0231, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.10481512546539307, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0229, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.1453929841518402, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0255, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.133439838886261, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0248, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.1380738615989685, + "learning_rate": 2.721037237880308e-05, + "loss": 0.023, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.1386265754699707, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0264, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.14298051595687866, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.0247, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.11756420135498047, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0222, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1197751984000206, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0216, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.2194000482559204, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0243, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.14157775044441223, + "learning_rate": 2.719594991888059e-05, + "loss": 0.023, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.14389759302139282, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0228, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.12442908436059952, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0267, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.1188371554017067, + "learning_rate": 2.718872618272836e-05, + "loss": 0.027, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11137005686759949, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0237, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.10907725989818573, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0251, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.150213360786438, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0229, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.2098330855369568, + "learning_rate": 2.717908157623383e-05, + "loss": 0.0245, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.12558938562870026, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0257, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.13659170269966125, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0247, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.12037219852209091, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0228, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.13085514307022095, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0252, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.13916544616222382, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0251, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.13098886609077454, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.021, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.12777045369148254, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0258, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.10204357653856277, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0249, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.11189272999763489, + "learning_rate": 2.715732712163776e-05, + "loss": 0.0232, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.1412421613931656, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0255, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.09470678865909576, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0217, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.09736761450767517, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0233, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12988215684890747, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.1080017015337944, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0263, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.12020932137966156, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0238, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.1058862954378128, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0247, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11686211824417114, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.024, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.11078432947397232, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0226, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.13394472002983093, + "learning_rate": 2.713306780076029e-05, + "loss": 0.0237, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.11299733072519302, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.025, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.1196199283003807, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0251, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.16477037966251373, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0241, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.1175335943698883, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.11024156212806702, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0237, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.12922869622707367, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0278, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.1429518610239029, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0254, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.14210145175457, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0225, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.12270870059728622, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0257, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.12436466664075851, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0258, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.10704488307237625, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0224, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.15890415012836456, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0252, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.12372975796461105, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0225, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.11652565747499466, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.028, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.1483478546142578, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0255, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.13188686966896057, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0238, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.18725869059562683, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0244, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.15949182212352753, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0225, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.14752954244613647, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0229, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.14210951328277588, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0233, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.1388787031173706, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.0253, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.11010867357254028, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0227, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.13336122035980225, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0227, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.10118173807859421, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.0265, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.1434611827135086, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0263, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.14133326709270477, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.026, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.15359432995319366, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0236, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.12927602231502533, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.022, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.10330940783023834, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0236, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.1264553815126419, + "learning_rate": 2.705973772304842e-05, + "loss": 0.025, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.15007789433002472, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0267, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.11090201884508133, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0257, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.12769529223442078, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0252, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.0977141410112381, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0244, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.14598220586776733, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0254, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.11231301724910736, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0235, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.09610529243946075, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0262, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.1549275815486908, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0234, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.11284032464027405, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.0217, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.14465086162090302, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.0235, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.13792403042316437, + "learning_rate": 2.703264323632037e-05, + "loss": 0.024, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.1438492238521576, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0239, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.12613703310489655, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0244, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.14072397351264954, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0225, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.1420842856168747, + "learning_rate": 2.702276326859174e-05, + "loss": 0.026, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.17447218298912048, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0243, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.1708984673023224, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0259, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.1555069535970688, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.0252, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.12473946064710617, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0227, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.15750949084758759, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0248, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.14009647071361542, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0267, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.1416911780834198, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0229, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.11944608390331268, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0235, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.126204252243042, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0268, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.1186814159154892, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0242, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.12699876725673676, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0239, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.10452742129564285, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0221, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.1422647386789322, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0247, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.11601094901561737, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0232, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12339281290769577, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0223, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1883407086133957, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0243, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.1730756312608719, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0294, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.10429747402667999, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0194, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.1358909010887146, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0217, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.14417894184589386, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0265, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.12299799919128418, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0236, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.1361788511276245, + "learning_rate": 2.696816263672279e-05, + "loss": 0.024, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.14492252469062805, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0241, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.1209225058555603, + "learning_rate": 2.696317709902802e-05, + "loss": 0.0239, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.11048539727926254, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.12423776835203171, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0243, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.1254356950521469, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.12912537157535553, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0254, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.10864657163619995, + "learning_rate": 2.695069735893089e-05, + "loss": 0.022, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.15519478917121887, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0253, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.15276901423931122, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0231, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.1401076465845108, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0271, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.14237190783023834, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.024, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1186559721827507, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0238, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.1254940629005432, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0233, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.1337885558605194, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0255, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12958411872386932, + "learning_rate": 2.69306825906724e-05, + "loss": 0.025, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.11610005795955658, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0238, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.13515442609786987, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0239, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1492868959903717, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0224, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.1293300986289978, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.021, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12354710698127747, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.13226614892482758, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.0247, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.1368253529071808, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0255, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.13044875860214233, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0228, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14107753336429596, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0221, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.13521252572536469, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0236, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.13718804717063904, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0217, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.15192179381847382, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.0247, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.13718710839748383, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0244, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11781664192676544, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0239, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.14431379735469818, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0248, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.15059512853622437, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0231, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.12076479196548462, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0242, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.19006308913230896, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0238, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.140592560172081, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0233, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.1528412401676178, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0229, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.16418038308620453, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0242, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.12185576558113098, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0254, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11594245582818985, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0248, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1216709092259407, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0234, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.12024594098329544, + "learning_rate": 2.686776313743295e-05, + "loss": 0.024, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.1321716010570526, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.0255, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.13785192370414734, + "learning_rate": 2.686270520700719e-05, + "loss": 0.025, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.12321551889181137, + "learning_rate": 2.6860174890357e-05, + "loss": 0.0233, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14606279134750366, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0249, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.17512187361717224, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0272, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.14974310994148254, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.0243, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.11758870631456375, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0241, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.11157143115997314, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0232, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.10899091511964798, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0231, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.1136092096567154, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0224, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.10306857526302338, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0211, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13241226971149445, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0222, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.11886361986398697, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0244, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.11962216347455978, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0237, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.14516247808933258, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.14592237770557404, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0253, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1278424859046936, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0233, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.12451537698507309, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0218, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.11746378242969513, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0266, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.14618739485740662, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0226, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.1512313336133957, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0233, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.14035750925540924, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0264, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.13232699036598206, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.023, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.14616237580776215, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.0233, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.12882372736930847, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0265, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.13778620958328247, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.024, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.15017054975032806, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0239, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.14781621098518372, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0215, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1670294851064682, + "learning_rate": 2.679407107451536e-05, + "loss": 0.022, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.1112075075507164, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0232, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.12219138443470001, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0202, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12922540307044983, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.11967732757329941, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0226, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.1377563774585724, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0224, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.14771930873394012, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0234, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.14800170063972473, + "learning_rate": 2.677617030053916e-05, + "loss": 0.0251, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.1233784481883049, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0232, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.16905057430267334, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0214, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.13751909136772156, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0248, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.13289029896259308, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0216, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.10788674652576447, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0216, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.13367260992527008, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1270209699869156, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0239, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.140412375330925, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0218, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.17788565158843994, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0253, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.13695330917835236, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.023, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.1250872015953064, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0232, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.11347977817058563, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0243, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.15546919405460358, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0233, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.13752058148384094, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.10431966185569763, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0215, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.17878347635269165, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0262, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.13218404352664948, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0241, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.12244930118322372, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.0247, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.11582386493682861, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0219, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.1160157322883606, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0203, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.10731764137744904, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0234, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.11152671277523041, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0218, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.12760375440120697, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0218, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.13706916570663452, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0219, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.14939993619918823, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0208, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.1169784739613533, + "learning_rate": 2.67092985767871e-05, + "loss": 0.0217, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.1569128930568695, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12233320623636246, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0251, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13152818381786346, + "learning_rate": 2.670154386126402e-05, + "loss": 0.0231, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.1529632955789566, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0246, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.11001774668693542, + "learning_rate": 2.669636960731296e-05, + "loss": 0.0223, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.1473267376422882, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.025, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.12853027880191803, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0246, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.15698745846748352, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0209, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12175270169973373, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0229, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.15634675323963165, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.0224, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.16060872375965118, + "learning_rate": 2.668082553346281e-05, + "loss": 0.0254, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.13773131370544434, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0224, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.14258436858654022, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0216, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.13032476603984833, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0215, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.17300327122211456, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0221, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.13187018036842346, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0226, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.14065369963645935, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0232, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.11484711617231369, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0263, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1380661576986313, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0209, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.17800134420394897, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0244, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.14836716651916504, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0277, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14563506841659546, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.023, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.11463280767202377, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.026, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.13092686235904694, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0222, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.12265385687351227, + "learning_rate": 2.664443191002912e-05, + "loss": 0.025, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.14121955633163452, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0242, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.12938837707042694, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0222, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.1922816038131714, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.0228, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.12596914172172546, + "learning_rate": 2.663400187275054e-05, + "loss": 0.022, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.14095240831375122, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0244, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.1080891340970993, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0228, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.13916714489459991, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.0239, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.1489475965499878, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0251, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.11781706660985947, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0207, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.11544635146856308, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0232, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.14827974140644073, + "learning_rate": 2.661571529941175e-05, + "loss": 0.0236, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.12821446359157562, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.0186, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.1303764134645462, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0234, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.1049366369843483, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0232, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.15925322473049164, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0275, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.13151033222675323, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0213, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.09425143152475357, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0215, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.15108847618103027, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0228, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.1374080330133438, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0231, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.16061817109584808, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0253, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.1707434207201004, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0259, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.12051897495985031, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0217, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.09191743284463882, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0224, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.12547235190868378, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0245, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14491257071495056, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0228, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.11896242201328278, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0239, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.14253611862659454, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0233, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.12803201377391815, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.0226, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.12893305718898773, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0231, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.16127154231071472, + "learning_rate": 2.656586262343925e-05, + "loss": 0.0234, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.13307608664035797, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0258, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.11211924254894257, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.0219, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.1672789305448532, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0248, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.12878142297267914, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0231, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.11429315060377121, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0233, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.13469499349594116, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0254, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.12753085792064667, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0258, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.11875704675912857, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0234, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.13718342781066895, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0223, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.10762278735637665, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.021, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.11651597917079926, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0222, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.12953802943229675, + "learning_rate": 2.653421326308085e-05, + "loss": 0.0235, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.14820824563503265, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.023, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.16258545219898224, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0266, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.14097236096858978, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0224, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.1039188802242279, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0216, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15014079213142395, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0246, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12850438058376312, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.023, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.12245986610651016, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0209, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.15426407754421234, + "learning_rate": 2.651304359235998e-05, + "loss": 0.025, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.13411583006381989, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0229, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.13705122470855713, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0254, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.13120216131210327, + "learning_rate": 2.650509053370553e-05, + "loss": 0.022, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14599482715129852, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0249, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.11628340184688568, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0215, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.13643144071102142, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0228, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.12846967577934265, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0214, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.13181492686271667, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0245, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1277213841676712, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0244, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12288910150527954, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0226, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.14789454638957977, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0215, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.11457780748605728, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0248, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.12511487305164337, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0206, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.12255935370922089, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0245, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.13214053213596344, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0238, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.1601077914237976, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.027, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.1244364008307457, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0212, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.11869437992572784, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0239, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.14327670633792877, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0227, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.12729141116142273, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0248, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.10953658074140549, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0205, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.20326738059520721, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0225, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.15614672005176544, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0237, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.11863692849874496, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0208, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.10683806240558624, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0224, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.1529647260904312, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.024, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.18428578972816467, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0232, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.19935141503810883, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.0223, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.14675194025039673, + "learning_rate": 2.643583521046136e-05, + "loss": 0.0216, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.21206484735012054, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0243, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.14521193504333496, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0214, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.11565426737070084, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0236, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.1087750792503357, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0215, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.11942578107118607, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0229, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.138235405087471, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0219, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.1079302579164505, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0222, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.15928085148334503, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0225, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.10920153558254242, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.024, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.12247785925865173, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0205, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.10454121232032776, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0238, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.1726839691400528, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0211, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1617811918258667, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0225, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.15452679991722107, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0214, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1573309749364853, + "learning_rate": 2.639561293166059e-05, + "loss": 0.022, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.13596323132514954, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0234, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.09188699722290039, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0219, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.12158972024917603, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0202, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.16615170240402222, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.0221, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.1163380891084671, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0255, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.17704762518405914, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0243, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.1591942012310028, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.021, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.14697043597698212, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0241, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1234351247549057, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0261, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.1379440575838089, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0266, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1025586649775505, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.022, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.11894206702709198, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0226, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1431341916322708, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0241, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.10291697829961777, + "learning_rate": 2.635789644394928e-05, + "loss": 0.022, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.1622016727924347, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0229, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.11867184937000275, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0262, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.18177814781665802, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0237, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.15159954130649567, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0218, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.11481288820505142, + "learning_rate": 2.634438526911474e-05, + "loss": 0.024, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.10989677160978317, + "learning_rate": 2.634168044873305e-05, + "loss": 0.024, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.12370797991752625, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0235, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1390318125486374, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0231, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.14030151069164276, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.0207, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11659140884876251, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.021, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.16727864742279053, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0242, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.11524021625518799, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0265, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.10776443779468536, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0224, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.16366012394428253, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0264, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.1102370023727417, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0245, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11543556302785873, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0216, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.11101235449314117, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0208, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.1257167011499405, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.024, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.1063276082277298, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0241, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.14303457736968994, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0242, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.12862786650657654, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0244, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15487778186798096, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.14244592189788818, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0211, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1066693440079689, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0199, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11721482872962952, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0214, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.13803230226039886, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.0227, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.11225558817386627, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.028, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.11660541594028473, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0239, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12551093101501465, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0221, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.1305139809846878, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.1281510442495346, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0235, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.14735114574432373, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0213, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.10699455440044403, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0231, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.16579580307006836, + "learning_rate": 2.626559653214893e-05, + "loss": 0.0213, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.09389153122901917, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0218, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.17565062642097473, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0245, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.1368194818496704, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0214, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.10599412024021149, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0221, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.12645669281482697, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0233, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.12325828522443771, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.0228, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.11353788524866104, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.10889376699924469, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0227, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.12935040891170502, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0214, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.1422247439622879, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0219, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.16723784804344177, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0239, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.09775881469249725, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.0229, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.13455381989479065, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0253, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.13700921833515167, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0239, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12252738326787949, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0231, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.15983854234218597, + "learning_rate": 2.622181856488204e-05, + "loss": 0.0223, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.18716242909431458, + "learning_rate": 2.621907518918971e-05, + "loss": 0.025, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.16380536556243896, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0229, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.1524866819381714, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0221, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.12281060218811035, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0213, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.09793791174888611, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0249, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.12271476536989212, + "learning_rate": 2.620534553467634e-05, + "loss": 0.0221, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.12616175413131714, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0236, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.10274055600166321, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0205, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.16873067617416382, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.027, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.10157257318496704, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0236, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.11863094568252563, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0225, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.09981630742549896, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0233, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11045375466346741, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0226, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.13111108541488647, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0232, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.12417030334472656, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0226, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.12029916048049927, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0217, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.10766672343015671, + "learning_rate": 2.617506544598029e-05, + "loss": 0.025, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.12039721757173538, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0228, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.11044551432132721, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0223, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.12463348358869553, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.0232, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.14382848143577576, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0238, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.13677646219730377, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0239, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.11536712944507599, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0224, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.12461430579423904, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0253, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.16901540756225586, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0242, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.12255802005529404, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0212, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.14185196161270142, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.0218, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.1292722374200821, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0249, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12001507729291916, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0222, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.12286248803138733, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0211, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.12487491965293884, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0231, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13417202234268188, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0257, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.11690660566091537, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0219, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15474916994571686, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0223, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.1720103919506073, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0239, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.16786889731884003, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0216, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.2279045283794403, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.0254, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13872499763965607, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0238, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.1881643831729889, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0211, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.1372661143541336, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0237, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.10514581203460693, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0217, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.11103228479623795, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0248, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1349911093711853, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0237, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.12774604558944702, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0245, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.15125125646591187, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0233, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.1190832257270813, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0248, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.1088283434510231, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0187, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.12167634069919586, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.023, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.15135347843170166, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0226, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.12857669591903687, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.0195, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.13588359951972961, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0234, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.14902308583259583, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0235, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.15095068514347076, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0209, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.1116727739572525, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0255, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11652971059083939, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0211, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.11875079572200775, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0223, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.11100098490715027, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.022, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.13648055493831635, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0234, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.12870723009109497, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0205, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.10456091910600662, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0232, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.13770641386508942, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0212, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.1526155173778534, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0244, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12664297223091125, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0241, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.12000498175621033, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0239, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.11930232495069504, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.0207, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.11384909600019455, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0217, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.1746685951948166, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.023, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.15868088603019714, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0233, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.15131928026676178, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0194, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.10279352962970734, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0215, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.13810355961322784, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0202, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.13024795055389404, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0217, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1374605894088745, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0215, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.14356477558612823, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0206, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.11046172678470612, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0236, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.10146678239107132, + "learning_rate": 2.601090850395125e-05, + "loss": 0.021, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13247323036193848, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0217, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.13672757148742676, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0201, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.08651759475469589, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.0196, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.14821568131446838, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0223, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11985234171152115, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0208, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11231720447540283, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0219, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.12338588386774063, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0202, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.15079368650913239, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0229, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.1348753422498703, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0209, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1333470344543457, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0207, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.10075798630714417, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0232, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1217522993683815, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0247, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.12041056901216507, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.022, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.10997799783945084, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0226, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.119750015437603, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0239, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.15246307849884033, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0239, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.1431925892829895, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0227, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.1181643158197403, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0227, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.13645906746387482, + "learning_rate": 2.595742267991401e-05, + "loss": 0.0213, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.12661530077457428, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0233, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.1344967484474182, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.0221, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.13433857262134552, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0238, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.10517163574695587, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0237, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.10689525306224823, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.022, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.13307900726795197, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0209, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.14696283638477325, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0261, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.12473653256893158, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0213, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.106630340218544, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0221, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.12092334777116776, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0232, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.11033511906862259, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0193, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11916322261095047, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0212, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.1204444020986557, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0228, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.1373489499092102, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0225, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.13124211132526398, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0208, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.18858365714550018, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0258, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.1558879315853119, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0228, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.1481737494468689, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0242, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.10424906760454178, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.021, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.15494640171527863, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0238, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.1801268309354782, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0238, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.14176428318023682, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0228, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.15850430727005005, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0235, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.14727924764156342, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0232, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.14907293021678925, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0242, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.11011895537376404, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0231, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.13541805744171143, + "learning_rate": 2.588090010443142e-05, + "loss": 0.0224, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.13078846037387848, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0228, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.14508767426013947, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0207, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.12445177882909775, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0227, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.14860141277313232, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0244, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.13311992585659027, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0207, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.1157015860080719, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0209, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.13310302793979645, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.0234, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.14026974141597748, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0246, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.14788632094860077, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0213, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.1551811844110489, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0241, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.10706475377082825, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.021, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.12536053359508514, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0228, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.1260751187801361, + "learning_rate": 2.584384079351628e-05, + "loss": 0.022, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.1082935482263565, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0207, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.13922622799873352, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0251, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1240227147936821, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0223, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.14512665569782257, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0218, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12859801948070526, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.0208, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.13418391346931458, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.021, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.10546918958425522, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.022, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.10668392479419708, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10407276451587677, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.02, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.13980497419834137, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0235, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.16233955323696136, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0224, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.16743852198123932, + "learning_rate": 2.580950866038058e-05, + "loss": 0.022, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.09152300655841827, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0206, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.14319349825382233, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0205, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.1045396700501442, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0219, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.10865644365549088, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0201, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.12166012078523636, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0217, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.12505589425563812, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0224, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.1353408247232437, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0242, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.1379590630531311, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0221, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12337373197078705, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0235, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.12092723697423935, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0233, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.09944337606430054, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0215, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.15161453187465668, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0193, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.0963459238409996, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0225, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.15534570813179016, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0247, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.142191082239151, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0228, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.1454230695962906, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0226, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.12661893665790558, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0211, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.11373333632946014, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0213, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12031935155391693, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.0208, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.13401862978935242, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0224, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.11285468935966492, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0231, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.12453637272119522, + "learning_rate": 2.574625966768285e-05, + "loss": 0.0211, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.13041555881500244, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1391122043132782, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0224, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.1345762014389038, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0244, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.12814867496490479, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0219, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.11058632284402847, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0191, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.15668845176696777, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.021, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.1509203314781189, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.0254, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.13209059834480286, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0201, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.1357070952653885, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0226, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.0896834284067154, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0207, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.1570383608341217, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0234, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.16916486620903015, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0254, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.1385425180196762, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.0218, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.11236828565597534, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0214, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11224918812513351, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0233, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.1305687576532364, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0219, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.15051157772541046, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0226, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.13418416678905487, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0244, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.14324583113193512, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0235, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.10810746997594833, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0197, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.14513953030109406, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0215, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1187531128525734, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0193, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11865831911563873, + "learning_rate": 2.567971342427169e-05, + "loss": 0.021, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.12040061503648758, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0229, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.10597001016139984, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0209, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.1219123974442482, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0242, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.11732123047113419, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.0203, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.15642790496349335, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0226, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.1893257051706314, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0247, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11829031258821487, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0239, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.14971555769443512, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0253, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.12424097210168839, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0251, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1071329414844513, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0271, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.11146610230207443, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0214, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.10925749689340591, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.0212, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.14797845482826233, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0221, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.10348086059093475, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0218, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.14607182145118713, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0217, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.11097954958677292, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0259, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.11300431936979294, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0197, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.11342521011829376, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0202, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.12889035046100616, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0217, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.10449834913015366, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0216, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1293669193983078, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0227, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.10667148977518082, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.021, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.0839434340596199, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0236, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.12906265258789062, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0223, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.1100582480430603, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0243, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.10644520074129105, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.0215, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.11693669110536575, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.022, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.13665053248405457, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0227, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12752027809619904, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0241, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.11534811556339264, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0234, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.1467696875333786, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0245, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10596702992916107, + "learning_rate": 2.558641420481684e-05, + "loss": 0.02, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.14685864746570587, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.022, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.13554386794567108, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.026, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.12572740018367767, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0232, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.1583555042743683, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0209, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.12913484871387482, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0226, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.11418721824884415, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0225, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.12722279131412506, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.0211, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.1382816582918167, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.0243, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.11612103134393692, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0212, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.1537163406610489, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.021, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.10462892800569534, + "learning_rate": 2.555415209239832e-05, + "loss": 0.0214, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.12759479880332947, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0232, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.1290607452392578, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0222, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12458004057407379, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0213, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.1113182008266449, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0223, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.15024979412555695, + "learning_rate": 2.553945541728904e-05, + "loss": 0.022, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.12657828629016876, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.025, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.13719463348388672, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.024, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.13465221226215363, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0254, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.08891251683235168, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0215, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.08946283161640167, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0224, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.13241387903690338, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0229, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.13068708777427673, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0215, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.12115749716758728, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.025, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.12863430380821228, + "learning_rate": 2.551295099554365e-05, + "loss": 0.0202, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.1332341581583023, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0204, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.102240651845932, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0199, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.16963273286819458, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0226, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12840960919857025, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0225, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.09958021342754364, + "learning_rate": 2.549819836273345e-05, + "loss": 0.02, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.10737278312444687, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0226, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.11473958194255829, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0219, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.12188851088285446, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0216, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.1229034960269928, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0227, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.1316317766904831, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1230575442314148, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0218, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13473588228225708, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0221, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1293632537126541, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0223, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.11663107573986053, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0198, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.147003173828125, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0216, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.10070289671421051, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0204, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12855689227581024, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0222, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.10310211777687073, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0227, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.25780248641967773, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0204, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.11621664464473724, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0196, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.15240156650543213, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0232, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.1359490305185318, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0191, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.13072653114795685, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.0242, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.1296282559633255, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0214, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.12183726578950882, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0215, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.10602327436208725, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0222, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.12424091249704361, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0215, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.14287705719470978, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0217, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.10017845034599304, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0214, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.1265304833650589, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0214, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.15279516577720642, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0194, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.11116814613342285, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0197, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.14380785822868347, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0187, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.12522602081298828, + "learning_rate": 2.541224092500406e-05, + "loss": 0.023, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.12386767566204071, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.0197, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11278095096349716, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0205, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12119927257299423, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0206, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.14128951728343964, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0197, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.12576138973236084, + "learning_rate": 2.539735331078111e-05, + "loss": 0.024, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.13276934623718262, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0217, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.15223100781440735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0222, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.13918957114219666, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0211, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.10129881650209427, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0192, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.15048135817050934, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0191, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.11579657346010208, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.019, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.1882474422454834, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0241, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.12608706951141357, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.021, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.14131402969360352, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0234, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.11546603590250015, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0208, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.0989922359585762, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0219, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.13154740631580353, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0234, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.12450987845659256, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0219, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.15696841478347778, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0228, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12059786915779114, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0236, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.1058221235871315, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0234, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.11429761350154877, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0232, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.11841994524002075, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0193, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.12387857586145401, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0239, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.11453676968812943, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0224, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.15571972727775574, + "learning_rate": 2.533461009735874e-05, + "loss": 0.022, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.10241609811782837, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.022, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.11682361364364624, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0211, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.13623486459255219, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0215, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.12484022229909897, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0212, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.09105533361434937, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0223, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.1024160087108612, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0174, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.14108790457248688, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0215, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.1268249899148941, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0211, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12255556881427765, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0192, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.15904900431632996, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0259, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.1356613039970398, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.023, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.12410393357276917, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0223, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1291753053665161, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0221, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.109084852039814, + "learning_rate": 2.529258887162501e-05, + "loss": 0.0243, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.16595998406410217, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0201, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.14062269032001495, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0232, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.13198593258857727, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0214, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.10668856650590897, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0229, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.11266572028398514, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0233, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.12742944061756134, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0221, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.12420892715454102, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0215, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.1265600174665451, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0215, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.11914686858654022, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0222, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.12876294553279877, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0233, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.09800750762224197, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0219, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1273864060640335, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0238, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.13299106061458588, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0226, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.09396592527627945, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.12955208122730255, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0204, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.16461864113807678, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0211, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.11535970121622086, + "learning_rate": 2.524135719213194e-05, + "loss": 0.02, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.11860322207212448, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0212, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10968036204576492, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0207, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.12494095414876938, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.02, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.10398012399673462, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0205, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.12939494848251343, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0216, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.10996326059103012, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0199, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.13704033195972443, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0208, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.11454308778047562, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.0212, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.14186057448387146, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0226, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.12624409794807434, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0199, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.14392296969890594, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.15774424374103546, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0213, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.1352480947971344, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0192, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.10594955086708069, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0231, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.11280091851949692, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0202, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11683343350887299, + "learning_rate": 2.519293379145164e-05, + "loss": 0.022, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.11721377819776535, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0223, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.18298691511154175, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.0225, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12927964329719543, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0202, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.1413726508617401, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.021, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.1330733746290207, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0228, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12341104447841644, + "learning_rate": 2.517472387591168e-05, + "loss": 0.021, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.12199880182743073, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.0202, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.12259527295827866, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.0227, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.11682213097810745, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0246, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.12783272564411163, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0204, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.09663784503936768, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0228, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.13686567544937134, + "learning_rate": 2.515648614306516e-05, + "loss": 0.02, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.12164216488599777, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0234, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.1371183544397354, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0217, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.12031334638595581, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0203, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.1183433011174202, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.1265946477651596, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0224, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.12686678767204285, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0202, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.13161398470401764, + "learning_rate": 2.513517369661726e-05, + "loss": 0.0216, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.13780610263347626, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0205, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.14213374257087708, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0237, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.12532930076122284, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0215, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1276065707206726, + "learning_rate": 2.512297821731217e-05, + "loss": 0.02, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.08893107622861862, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12819649279117584, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0202, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.09707193821668625, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0202, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.511077043763402e-05, + "loss": 0.0208, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.10047636926174164, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0219, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.15637417137622833, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0218, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.15387912094593048, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0213, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.14876650273799896, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0222, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.13631325960159302, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0214, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.12577500939369202, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0232, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15896311402320862, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0237, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.16459719836711884, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0223, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11081109195947647, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0223, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.12327340245246887, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0219, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14812195301055908, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0228, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.11862681806087494, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0252, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.10684223473072052, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0227, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.1279568374156952, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0216, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.13341103494167328, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0213, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.11913862824440002, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0214, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.13493330776691437, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.0214, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.1272144317626953, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0209, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.11348695307970047, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0217, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.16526950895786285, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0198, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.14953579008579254, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0231, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.12980960309505463, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0226, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.16652929782867432, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0244, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1722235530614853, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.026, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.23629626631736755, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0221, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.12417787313461304, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.023, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.11768155544996262, + "learning_rate": 2.502804732645032e-05, + "loss": 0.0186, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.13731959462165833, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.023, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.18034201860427856, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.0201, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11240512877702713, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0217, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.10463474690914154, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0201, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.1273874044418335, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0194, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.12314468622207642, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0237, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.13170552253723145, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.021, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.12725704908370972, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0198, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.11761748790740967, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0198, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13019445538520813, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0237, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.09089519083499908, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.0202, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.13669925928115845, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0236, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.0891583263874054, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0189, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.12015428394079208, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.0214, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.12889859080314636, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0214, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.1092396154999733, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.022, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.1011805608868599, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0201, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.11652060598134995, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.019, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.12280789762735367, + "learning_rate": 2.496950169240971e-05, + "loss": 0.0229, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.1253744661808014, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0245, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.15854160487651825, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0257, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.11671199649572372, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0204, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.15195102989673615, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.11878954619169235, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0207, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.1035882979631424, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0195, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12105780094861984, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0191, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.13278989493846893, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0207, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.16576215624809265, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.0228, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.13104894757270813, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0197, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.14476071298122406, + "learning_rate": 2.49354818242132e-05, + "loss": 0.023, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.14116041362285614, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0204, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.133507639169693, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0213, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11147522926330566, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0192, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.1317354440689087, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0229, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.15885432064533234, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0214, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.0996033102273941, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0213, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.11059842258691788, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0181, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.1103382259607315, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0252, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.12147006392478943, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.1167205348610878, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0217, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.11838674545288086, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0223, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.1745499074459076, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0213, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.09892687201499939, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0213, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.17135944962501526, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0228, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.13930797576904297, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0211, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.10764696449041367, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0204, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11211943626403809, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0207, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1359340399503708, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0215, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.11819283664226532, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0213, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.1166793555021286, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0216, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.14411170780658722, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0206, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.1380920261144638, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0211, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.12901239097118378, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.0224, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.1171313151717186, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.0207, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.15169396996498108, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0225, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.2066943198442459, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0182, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1179543286561966, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0226, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.12273375689983368, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.0224, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.09459412097930908, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0177, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.10208278894424438, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0207, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.11947600543498993, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.0246, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10635806620121002, + "learning_rate": 2.483599695058918e-05, + "loss": 0.021, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.15165211260318756, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0242, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.10573819279670715, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0221, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.10153646767139435, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0212, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.12596352398395538, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.021, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.11259406059980392, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0214, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.09709548950195312, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0186, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1462256908416748, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0228, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.15480898320674896, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0245, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.10707002878189087, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0197, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.10963082313537598, + "learning_rate": 2.480475069778109e-05, + "loss": 0.019, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.13645771145820618, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0229, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.08208916336297989, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0194, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.11223483830690384, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0212, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.12233787775039673, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0228, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.12175288051366806, + "learning_rate": 2.478909963381241e-05, + "loss": 0.0204, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.14267505705356598, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0203, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.12649142742156982, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0217, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.11254285275936127, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0214, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.1454164832830429, + "learning_rate": 2.477656539969673e-05, + "loss": 0.0221, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.17177407443523407, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0212, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.1089746281504631, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0222, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.1525813490152359, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0236, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.09651731699705124, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0214, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.12710145115852356, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0233, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.15999987721443176, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0215, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1600969433784485, + "learning_rate": 2.475460191424065e-05, + "loss": 0.0201, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.11045356094837189, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0208, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.14192573726177216, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.0216, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.11450321227312088, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0202, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.13454747200012207, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.021, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.11518846452236176, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0194, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.10673099011182785, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0214, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.12131883949041367, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0203, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.18098120391368866, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0182, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.09981405735015869, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0214, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.09194957464933395, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0209, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.11133445054292679, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0227, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.14570456743240356, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0181, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.14878204464912415, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0213, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.1475936472415924, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0229, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.14091694355010986, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0229, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.10411489754915237, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.021, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.10015816241502762, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0211, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09339930862188339, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0206, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.13016946613788605, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0223, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.10682770609855652, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0258, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.13658446073532104, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0232, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.12742926180362701, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.13928014039993286, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0208, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.10824346542358398, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0199, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.11346063017845154, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.0225, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12237603217363358, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0226, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.12201637774705887, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0209, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.09991955012083054, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0239, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.15387621521949768, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.025, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.1588744819164276, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0192, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.1347019523382187, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0196, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.12746664881706238, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0212, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11522521078586578, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0197, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.09697534143924713, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0219, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12183485180139542, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0195, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.13119301199913025, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0195, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.10748504102230072, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0192, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.1617487072944641, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0224, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.11518653482198715, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0207, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.17646238207817078, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0233, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.13729365170001984, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.09777814894914627, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0207, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.117672860622406, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0254, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.09014946967363358, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0214, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.1273731142282486, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0223, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10932891070842743, + "learning_rate": 2.460937119575347e-05, + "loss": 0.0224, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.1672651767730713, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0205, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.0930962786078453, + "learning_rate": 2.46030216185615e-05, + "loss": 0.02, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.09708375483751297, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0192, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.13240690529346466, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0214, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.10310725122690201, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0204, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.13352994620800018, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0206, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.15277482569217682, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0222, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.14223086833953857, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0229, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12845414876937866, + "learning_rate": 2.458077513854331e-05, + "loss": 0.0221, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13195961713790894, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0232, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.1461530476808548, + "learning_rate": 2.457441244940383e-05, + "loss": 0.022, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13729079067707062, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0196, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.15615087747573853, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0213, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.12907880544662476, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0198, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.11320316791534424, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0225, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.14982816576957703, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0225, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.12824060022830963, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0196, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.11940523982048035, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0204, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946357995271683, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.0218, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.13296495378017426, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0229, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.1383914053440094, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0207, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.11321834474802017, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0202, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.1276783049106598, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0193, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1406320035457611, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0191, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.12060298770666122, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0195, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.08627335727214813, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0191, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1429424285888672, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0197, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.15133187174797058, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0212, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.11892743408679962, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0231, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.11871744692325592, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0211, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.105750173330307, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.0234, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10969540476799011, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.022, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.09607745707035065, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0173, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.11796857416629791, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0193, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.114372618496418, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.02, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.1164821982383728, + "learning_rate": 2.449463406055895e-05, + "loss": 0.023, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.13561253249645233, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.018, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.14956215023994446, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0245, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.12785622477531433, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0222, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.16085495054721832, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.021, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10792174935340881, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0209, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.10508404672145844, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0217, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.11471311002969742, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0202, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.14265185594558716, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0224, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.13569262623786926, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0196, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15635624527931213, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0205, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12042741477489471, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0179, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.09488346427679062, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0198, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.20503900945186615, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0199, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.13667535781860352, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13254515826702118, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0204, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.13879507780075073, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0233, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.11448678374290466, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0237, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1255042403936386, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0204, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11896912008523941, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0207, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.12624827027320862, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0194, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.09312308579683304, + "learning_rate": 2.442727178914008e-05, + "loss": 0.022, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12632645666599274, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0189, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.11468527466058731, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0182, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.09245705604553223, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0188, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.12113099545240402, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0191, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.13049040734767914, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0221, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.16979938745498657, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0211, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.11175577342510223, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0172, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.1191241443157196, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0206, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.16317729651927948, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.0213, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.10087717324495316, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0194, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.12195733189582825, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0253, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11880146712064743, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0222, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.14213789999485016, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0198, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.15624478459358215, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.0205, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09695806354284286, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0219, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11260177940130234, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0206, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.15394112467765808, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0219, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.11663617193698883, + "learning_rate": 2.436928131964383e-05, + "loss": 0.0214, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.12596042454242706, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0225, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.10310690104961395, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0203, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.12332364916801453, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0203, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.16257967054843903, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0234, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.13102053105831146, + "learning_rate": 2.435313188161029e-05, + "loss": 0.022, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.14736537635326385, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0199, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.14611469209194183, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0238, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.14594200253486633, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0212, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.16921328008174896, + "learning_rate": 2.434019954415699e-05, + "loss": 0.0236, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.1318638175725937, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0201, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.10869060456752777, + "learning_rate": 2.433372911848899e-05, + "loss": 0.019, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.1446760594844818, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.021, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.1298162341117859, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10307328402996063, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0197, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.11932921409606934, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0209, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.10986137390136719, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0227, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.12099824845790863, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0182, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.13432320952415466, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.0213, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.14018790423870087, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0199, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.13015322387218475, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0191, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.10281360149383545, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0215, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.09500012546777725, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0216, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.15131708979606628, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0225, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.08750247210264206, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0221, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.10143480449914932, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0214, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.12935201823711395, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0202, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.10703578591346741, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0207, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08649012446403503, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0184, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.11186458170413971, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.021, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.11181459575891495, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0211, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.11756999045610428, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0216, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.12061341106891632, + "learning_rate": 2.426561884159603e-05, + "loss": 0.019, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.13861820101737976, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0214, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.11545911431312561, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0195, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.14892302453517914, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0226, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.10354511439800262, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0197, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.15290679037570953, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0208, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.12185964733362198, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0205, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.10486926138401031, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0206, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.10959216207265854, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0198, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.11638635396957397, + "learning_rate": 2.423633359536448e-05, + "loss": 0.0227, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.13661672174930573, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0206, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.13071168959140778, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0189, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.16377446055412292, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0203, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.22221329808235168, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0185, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.11251460015773773, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0191, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.10873523354530334, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0228, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.13705629110336304, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0203, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.13980528712272644, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.10355913639068604, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0199, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.1282302737236023, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.0215, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.1221374049782753, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0192, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.18542620539665222, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0199, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1250954121351242, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0189, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12665164470672607, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0217, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.09464617818593979, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.0197, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.09657798707485199, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0194, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.11533467471599579, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0189, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.13647402822971344, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0194, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.13276426494121552, + "learning_rate": 2.41743228236291e-05, + "loss": 0.022, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.16899874806404114, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0209, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.14313359558582306, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0211, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.14507170021533966, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0218, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.13025349378585815, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0209, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.1666184365749359, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0217, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14891882240772247, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0205, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11392978578805923, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0223, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.12964752316474915, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.019, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11125730723142624, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0211, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.08378217369318008, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0203, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.14530996978282928, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0206, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.10187243670225143, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0188, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.09652446955442429, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0192, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.10552513599395752, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0212, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11044801771640778, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0215, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.12248426675796509, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0185, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.11164992302656174, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0198, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.11919678002595901, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0187, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1675145924091339, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0262, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.15451745688915253, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0226, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.0997307077050209, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0196, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.1272345781326294, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0214, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.1286081224679947, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0214, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12226564437150955, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0211, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.11795010417699814, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0216, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.12963736057281494, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0215, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.10415369272232056, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0215, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.11642655730247498, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0194, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.19006073474884033, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0216, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.16677096486091614, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.0235, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.15928123891353607, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0195, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.10270075500011444, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0197, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.12253808975219727, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0216, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14145877957344055, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0211, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1387409120798111, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0185, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.11433693021535873, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0204, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.14247572422027588, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0187, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.12203793972730637, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.018, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.15705470740795135, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0212, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.10648002475500107, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0182, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.11573450267314911, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0211, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.11960671842098236, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0245, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.14416365325450897, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0178, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.11234421283006668, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0205, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.1375260204076767, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.025, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.08686859905719757, + "learning_rate": 2.402315322351119e-05, + "loss": 0.018, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11352542787790298, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0198, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.13374295830726624, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0216, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.09358088672161102, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0202, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.13767337799072266, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0229, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.1405765861272812, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.02, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11317892372608185, + "learning_rate": 2.400332803040664e-05, + "loss": 0.02, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.12356667965650558, + "learning_rate": 2.400002143698994e-05, + "loss": 0.0225, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.12570610642433167, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0232, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.1461946964263916, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0201, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.10893318802118301, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0211, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.10969080775976181, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0235, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.17320629954338074, + "learning_rate": 2.398347822254704e-05, + "loss": 0.0221, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14296430349349976, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0206, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12349261343479156, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0182, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.11617975682020187, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0209, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.1574973613023758, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0222, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.157145157456398, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.0199, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.14698748290538788, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0203, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.12345635890960693, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0212, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.1370248794555664, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0194, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.1332298070192337, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0219, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.12525111436843872, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.0219, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.14905311167240143, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0211, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12063094973564148, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.0262, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.10361988842487335, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0195, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.09919667989015579, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0201, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.11141914129257202, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.0223, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.17785966396331787, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0201, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.126350998878479, + "learning_rate": 2.392710390356112e-05, + "loss": 0.0201, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.12107345461845398, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0213, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.11466097086668015, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0202, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.15397872030735016, + "learning_rate": 2.391713512040237e-05, + "loss": 0.022, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.14412778615951538, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0223, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.10495682060718536, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0194, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.13021183013916016, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0199, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.13993357121944427, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0182, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.128571018576622, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.0181, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.11090676486492157, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0204, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.10988310724496841, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0203, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.11705116927623749, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0214, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.1277252584695816, + "learning_rate": 2.388719222953459e-05, + "loss": 0.0225, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1561414748430252, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0221, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.15278629958629608, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.02, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14865094423294067, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0213, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.1127776950597763, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0211, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.15085943043231964, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0212, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14555132389068604, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0216, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.14708220958709717, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0209, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.1093590185046196, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0199, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.1303861141204834, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0242, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.11462346464395523, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0197, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.12198365479707718, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0206, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.11193065345287323, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0201, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.13254034519195557, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0181, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.13355755805969238, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0204, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12551282346248627, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0212, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1441735178232193, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0217, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.11511780321598053, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0218, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.10427968204021454, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0222, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.11364301294088364, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0166, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.1401003897190094, + "learning_rate": 2.38204570134882e-05, + "loss": 0.0215, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.11645112931728363, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0198, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1340353935956955, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0212, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.0982050895690918, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0208, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.16713382303714752, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0185, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.13803045451641083, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.021, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.15552546083927155, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0233, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.10836438834667206, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0211, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.1831241250038147, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0198, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.12369874119758606, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0193, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.10911241918802261, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0221, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.11441849172115326, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0236, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.11105310916900635, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0206, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.13617953658103943, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0189, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.11214766651391983, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0192, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.12536488473415375, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0206, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.10356712341308594, + "learning_rate": 2.376687582517292e-05, + "loss": 0.019, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.13332195580005646, + "learning_rate": 2.376352133090187e-05, + "loss": 0.0216, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.13996818661689758, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0216, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.14305290579795837, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0204, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.10494769364595413, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0202, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.1461218297481537, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.0197, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.12469783425331116, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0207, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.13164709508419037, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.0207, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.1553804576396942, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0228, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1445176750421524, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0217, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12673451006412506, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0164, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.17502479255199432, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0208, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.11328406631946564, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0196, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12234057486057281, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0193, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.1333971917629242, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0214, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.11227905005216599, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0183, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.1287100911140442, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0198, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.10215240716934204, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0194, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.09786848723888397, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0202, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.11103200167417526, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0219, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.1491294801235199, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0202, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.1364760547876358, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.0198, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.11583423614501953, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0206, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.13896186649799347, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0183, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.0970211774110794, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.019, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.10326618701219559, + "learning_rate": 2.368281439559657e-05, + "loss": 0.022, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.11581890285015106, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0234, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10186298936605453, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0205, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11655393242835999, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0193, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.11990638077259064, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0217, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.10573728382587433, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0177, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.09257248789072037, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.0216, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.11310117691755295, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0188, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.12632012367248535, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0217, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1535339206457138, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0209, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.13133156299591064, + "learning_rate": 2.364907420469738e-05, + "loss": 0.0207, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11225332319736481, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0217, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.10744500905275345, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0212, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.10509403049945831, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0208, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.16215650737285614, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0199, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1208849623799324, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0196, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.14882729947566986, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0229, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.18775896728038788, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0206, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.14199234545230865, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0201, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.1510499119758606, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.021, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.16389605402946472, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0197, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.13032132387161255, + "learning_rate": 2.361188413960337e-05, + "loss": 0.0211, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.13895465433597565, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0222, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.10914628207683563, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0206, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.12215261161327362, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0186, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12954379618167877, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.0193, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.12779726088047028, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.02, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.11825352907180786, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0214, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11001764982938766, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0211, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.10464788973331451, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.02, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.0892137810587883, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0198, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.09811265021562576, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0191, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.14199955761432648, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0196, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.13846102356910706, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0216, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.12102426588535309, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0214, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.12360251694917679, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0199, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.12587665021419525, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.15562282502651215, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.0214, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.11366713047027588, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0192, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.15177540481090546, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0204, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.08908293396234512, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0186, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.1349143236875534, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0197, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.09388674795627594, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0205, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.11973906308412552, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0198, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.1290522664785385, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.02, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.12087979167699814, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0207, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.10237318277359009, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0189, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.10426577180624008, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0204, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13028934597969055, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0197, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.10696206241846085, + "learning_rate": 2.351686222335574e-05, + "loss": 0.0206, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.11024264991283417, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0189, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.11294405907392502, + "learning_rate": 2.35100554725415e-05, + "loss": 0.0228, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.1188240721821785, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0204, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.17485342919826508, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0202, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12582673132419586, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0209, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.18180538713932037, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0202, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.1459868997335434, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0189, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.09818759560585022, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0195, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.1473611295223236, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0253, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.09137327969074249, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0209, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.14026956260204315, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0202, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.1511334776878357, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0182, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.15016451478004456, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0204, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.14407901465892792, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.0205, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.1332983374595642, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0192, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.14312349259853363, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.0195, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.11917808651924133, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0194, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.17896629869937897, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0203, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.12084568291902542, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0199, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.14310620725154877, + "learning_rate": 2.344867863274048e-05, + "loss": 0.0201, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1168353408575058, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0225, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.14718633890151978, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0204, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.14102597534656525, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0197, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.09543738514184952, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0218, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1074792668223381, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0198, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.11555791646242142, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0196, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.12645067274570465, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0191, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.11634501069784164, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.02, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.09201087802648544, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0204, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.1155405044555664, + "learning_rate": 2.341449046471072e-05, + "loss": 0.0215, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.09871169179677963, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0216, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.11332481354475021, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0179, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.12474322319030762, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.0177, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.11339092999696732, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0221, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.1200975552201271, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.0209, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.14456698298454285, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0211, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.1371302306652069, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0165, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.11727641522884369, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0232, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.11336934566497803, + "learning_rate": 2.338366646946928e-05, + "loss": 0.0202, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.10325363278388977, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0209, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.11268003284931183, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0187, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10260385274887085, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.0195, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.10920321941375732, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0225, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.09120934456586838, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0194, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11567433178424835, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0197, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.10144805908203125, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0206, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1246950775384903, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0215, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.100107841193676, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0197, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.15822021663188934, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0202, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13518433272838593, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.02, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12268074601888657, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0196, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.12333078682422638, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0184, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.13588106632232666, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0218, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.12120439857244492, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0197, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.16428445279598236, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0198, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.11615289747714996, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.0182, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.1432264745235443, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0207, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.11717039346694946, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0198, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.10115417838096619, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0193, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.11894490569829941, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0198, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.11940047889947891, + "learning_rate": 2.330810220141532e-05, + "loss": 0.021, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.10992413759231567, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0195, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.10544339567422867, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0182, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.11725642532110214, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.0198, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.09130900353193283, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0209, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.09033828228712082, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0177, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.11721840500831604, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0207, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.12251118570566177, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0215, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.1496249884366989, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0219, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.14808881282806396, + "learning_rate": 2.327710137160376e-05, + "loss": 0.0212, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.145681232213974, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.019, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.12777844071388245, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0173, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.1110587939620018, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0198, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.11348012834787369, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0207, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.11685867607593536, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.02, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.17935989797115326, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0211, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.12398742884397507, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0191, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.10836582630872726, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0196, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.12038073688745499, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0207, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.12211567908525467, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0206, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.1702142357826233, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0227, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12037540227174759, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0191, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.15964926779270172, + "learning_rate": 2.323223255665661e-05, + "loss": 0.018, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.12493535876274109, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0212, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.1359032839536667, + "learning_rate": 2.32253202698687e-05, + "loss": 0.0219, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.18753379583358765, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0228, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.1336824893951416, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0201, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.09901099652051926, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0221, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.12925978004932404, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0206, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.1278320550918579, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.0178, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.12977559864521027, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0176, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10098735243082047, + "learning_rate": 2.320110760314603e-05, + "loss": 0.023, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.12594632804393768, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.0209, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.1140967383980751, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0181, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.1287144273519516, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0195, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.12329071760177612, + "learning_rate": 2.318725808577064e-05, + "loss": 0.0198, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.1339070051908493, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10581240057945251, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0265, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.13716311752796173, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0213, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.14610108733177185, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0195, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.128855898976326, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0185, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.0950121283531189, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0223, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.14444753527641296, + "learning_rate": 2.316299750241968e-05, + "loss": 0.017, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.12585613131523132, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0177, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.1290416419506073, + "learning_rate": 2.315606032389525e-05, + "loss": 0.02, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.10740494728088379, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0191, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11578185111284256, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0211, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.11840540170669556, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0209, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.09531640261411667, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0202, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.10436122864484787, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0188, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.14102807641029358, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0205, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14070506393909454, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0186, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.09883972257375717, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0196, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11524099856615067, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0187, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.11240902543067932, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0176, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.08062270283699036, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.019, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.13965697586536407, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0202, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.11376950889825821, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0218, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.13540907204151154, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0188, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.10888075083494186, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0216, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.10280206799507141, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0178, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.11512117087841034, + "learning_rate": 2.309699448641602e-05, + "loss": 0.02, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.09106973558664322, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0182, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10531709343194962, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0204, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.159554585814476, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0173, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.12161346524953842, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0196, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.08224594593048096, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0175, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.12767191231250763, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0234, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.12864384055137634, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0224, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.10232329368591309, + "learning_rate": 2.3069137227667e-05, + "loss": 0.0208, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.09609836339950562, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0183, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.1301981806755066, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0202, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.1258714497089386, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0217, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.12543867528438568, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13326771557331085, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.022, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.11770334839820862, + "learning_rate": 2.304821853869955e-05, + "loss": 0.019, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.11729256063699722, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0187, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.11887779086828232, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0224, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.10921293497085571, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0192, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.1346455216407776, + "learning_rate": 2.303426051834282e-05, + "loss": 0.022, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.1177918091416359, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0209, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.12912803888320923, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0191, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.12571996450424194, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0211, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.13959148526191711, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0198, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.1404334306716919, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0211, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.1111592948436737, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0213, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.10322082787752151, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0203, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.11235317587852478, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0187, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.13599546253681183, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0205, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11067865788936615, + "learning_rate": 2.299932279415038e-05, + "loss": 0.02, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.14130094647407532, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0204, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.14370687305927277, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0199, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.13825304806232452, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0197, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.11885961890220642, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.02, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.12376029789447784, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0199, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.10103051364421844, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0181, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.10931621491909027, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0198, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.11835317313671112, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0202, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.08036374300718307, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0189, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.08135218918323517, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0196, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.12611854076385498, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0211, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.09217566251754761, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0181, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.10258878767490387, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0194, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.1165921688079834, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0196, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.09290144592523575, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.0205, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.11575523763895035, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0204, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13866478204727173, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0178, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.13036414980888367, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0191, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11062757670879364, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0193, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.09973529726266861, + "learning_rate": 2.292926536293724e-05, + "loss": 0.02, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.13707594573497772, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0186, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11635670065879822, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0185, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.13926011323928833, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0192, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.13966591656208038, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0194, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.09878037869930267, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0191, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.11935862898826599, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0193, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.15291833877563477, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0189, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.12976837158203125, + "learning_rate": 2.290117482755701e-05, + "loss": 0.017, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.13367557525634766, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0178, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12121888250112534, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0191, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.13135644793510437, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0226, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.1142699271440506, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.017, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.09670589119195938, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0199, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.12751951813697815, + "learning_rate": 2.288008171679078e-05, + "loss": 0.018, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.14463315904140472, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0184, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.11006469279527664, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0198, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.15361396968364716, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0186, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.20782725512981415, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0208, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.15324300527572632, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.018, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.17235179245471954, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0218, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.12993113696575165, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0198, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.1496446281671524, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0203, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.16271376609802246, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.0215, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11106697469949722, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0206, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.10508760809898376, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0218, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.11433881521224976, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0203, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.12877102196216583, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0203, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.13301214575767517, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0225, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.15351711213588715, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0213, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.1114102303981781, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0212, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.09978143125772476, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0198, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10524465143680573, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0183, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.09780002385377884, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0185, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.1361900269985199, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0218, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.1329333484172821, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0191, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.10501913726329803, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.019, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.1329827904701233, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0198, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.10849577933549881, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0213, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.14308017492294312, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.0199, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.12313494831323624, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0193, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.10307908058166504, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.0215, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10591195523738861, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0178, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.1411430686712265, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0237, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13116225600242615, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0203, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.12032114714384079, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0201, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.12519164383411407, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0177, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.10340812057256699, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0194, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.10876207798719406, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0207, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11245796829462051, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0207, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11444602906703949, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0171, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12232744693756104, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0196, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.10049490630626678, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0179, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.09973384439945221, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.022, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.11037951707839966, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0197, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.14473563432693481, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0207, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.1138916090130806, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0198, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.12471994757652283, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0182, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.12418543547391891, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0195, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.12200118601322174, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.0217, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.13389603793621063, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0194, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10176926106214523, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.0189, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.10195910185575485, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0202, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.13289214670658112, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.018, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.10307228565216064, + "learning_rate": 2.270347347405331e-05, + "loss": 0.0183, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.149918794631958, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0172, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.1182866096496582, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0173, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.1035924181342125, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0194, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.108797088265419, + "learning_rate": 2.268928118647676e-05, + "loss": 0.0174, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.11457139253616333, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0208, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.10256518423557281, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0188, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13053470849990845, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0209, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.11632310599088669, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0187, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.12769649922847748, + "learning_rate": 2.267152769015067e-05, + "loss": 0.0206, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.13292407989501953, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0191, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.09822595119476318, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0189, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.13729946315288544, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0229, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.15705017745494843, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0203, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.12355367839336395, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0201, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.1278308629989624, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0205, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.09874208271503448, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0218, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.1316995620727539, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0212, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.10361800342798233, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0209, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.09672107547521591, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0205, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.12609392404556274, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0205, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.1360992044210434, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0188, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.11627763509750366, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0206, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.11733578145503998, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.02, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.1705910563468933, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0174, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.11704754829406738, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0199, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.13247141242027283, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0224, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.10916013270616531, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0202, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.11092681437730789, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.0186, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.12811405956745148, + "learning_rate": 2.260036839187291e-05, + "loss": 0.02, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.11893392354249954, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0194, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.16553092002868652, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0189, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.12431131303310394, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.02, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10850611329078674, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0222, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.12647055089473724, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0193, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.10432784259319305, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0196, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.1159503236413002, + "learning_rate": 2.257540798155951e-05, + "loss": 0.0192, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.10592516511678696, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0219, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.1318269520998001, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0195, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.11932560801506042, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0194, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.16499531269073486, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.02, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.1334158480167389, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.0217, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.14781776070594788, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0206, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.14816460013389587, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0206, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.11319883167743683, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0196, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.12413972616195679, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0206, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.1277540624141693, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0211, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.12178509682416916, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0192, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.1210661232471466, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0232, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.11300696432590485, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0182, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.15876832604408264, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0194, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.1387646645307541, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0202, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.12502169609069824, + "learning_rate": 2.251824990920905e-05, + "loss": 0.02, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.12365051358938217, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0197, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.12610606849193573, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0194, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.13096576929092407, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0208, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.13002553582191467, + "learning_rate": 2.250393750922939e-05, + "loss": 0.02, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.10048994421958923, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0193, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.11463778465986252, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0179, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10637608915567398, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0182, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.11959703266620636, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0183, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.11852603405714035, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12075617164373398, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0186, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.09626025706529617, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0195, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12773670256137848, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.019, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13364142179489136, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.019, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.14562146365642548, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0207, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.11752697080373764, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.13970085978507996, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0219, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.12456551194190979, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0203, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.36701101064682007, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0191, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.17254851758480072, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0193, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.1397981345653534, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0205, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.1273665875196457, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0227, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.08652572333812714, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0194, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.118954136967659, + "learning_rate": 2.243582938893974e-05, + "loss": 0.019, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.12105695903301239, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0173, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.10059279948472977, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0189, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.1476186215877533, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0192, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.11198431998491287, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0212, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.14829197525978088, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0216, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.14479608833789825, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0225, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.16737256944179535, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0202, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.12101127207279205, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0198, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.09408389776945114, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.019, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.10649813711643219, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0191, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.1328127533197403, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.02, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1135721281170845, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0181, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.14464536309242249, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0167, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.15146523714065552, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0216, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11212945729494095, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0189, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.11722676455974579, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0185, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10031717270612717, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0181, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.14852827787399292, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.02, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12114712595939636, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0207, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.12325385212898254, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0185, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.12915199995040894, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0188, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.10423776507377625, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0195, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.13398106396198273, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0182, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.1426725685596466, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0214, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.12515439093112946, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0184, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.15888792276382446, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0183, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.11037078499794006, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0199, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.10706029832363129, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.0195, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.11116144061088562, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0202, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.12716875970363617, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0182, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.1446460634469986, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0172, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.09947994351387024, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0205, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.09251125156879425, + "learning_rate": 2.231705291760663e-05, + "loss": 0.018, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.08440345525741577, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0181, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12813445925712585, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0206, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.1017315611243248, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0192, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.13247326016426086, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0212, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.14845387637615204, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0194, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.13824601471424103, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0224, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.1151830404996872, + "learning_rate": 2.229177978101565e-05, + "loss": 0.0176, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.12420724332332611, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0179, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.13085328042507172, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0179, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.12436439841985703, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.11873050034046173, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0197, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.1338828206062317, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0212, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.1106492280960083, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0179, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.16542823612689972, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0202, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.08829429000616074, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0176, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.10792568325996399, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0186, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.1279839128255844, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0221, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.10742732882499695, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0217, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.10220900177955627, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0201, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.1660427749156952, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.019, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.13747647404670715, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0225, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.13351675868034363, + "learning_rate": 2.223753180973339e-05, + "loss": 0.021, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.14151513576507568, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.0195, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.12823058664798737, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0189, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.12779751420021057, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0177, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.2243196815252304, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0183, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.12365520745515823, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0188, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.10981941968202591, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0191, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.11569637805223465, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0195, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.15785852074623108, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0222, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11610189825296402, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0172, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.1124223843216896, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0184, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.14634257555007935, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.0196, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.13306891918182373, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0204, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.08198657631874084, + "learning_rate": 2.219041683306619e-05, + "loss": 0.0177, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.09529033303260803, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0166, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.11170824617147446, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0195, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13825464248657227, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0191, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.11621235311031342, + "learning_rate": 2.217590132288247e-05, + "loss": 0.0202, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11921709030866623, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.019, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12485305964946747, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0205, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.1244436725974083, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0204, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11240405589342117, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0197, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.1087997704744339, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0183, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.11403695493936539, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0172, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.12764202058315277, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0161, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.17418256402015686, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0171, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.1522650271654129, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0197, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.16773882508277893, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0214, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.14681261777877808, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0208, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.14842623472213745, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.021, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.10614728182554245, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0182, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.15053948760032654, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0209, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.1381484866142273, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.0176, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.13161014020442963, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0174, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.09695997089147568, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0202, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.08903522044420242, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0171, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11456362903118134, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0183, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.1260307878255844, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.02, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.11732659488916397, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0197, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.126072496175766, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0196, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.11586430668830872, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.02, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.10504680871963501, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0193, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.09652367234230042, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0186, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10370496660470963, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0176, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.1192321851849556, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0178, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.10527302324771881, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0182, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10162831097841263, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0191, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.16128556430339813, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0203, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.11897512525320053, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0204, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.14880484342575073, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0205, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.11369120329618454, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.0193, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.11618658900260925, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0176, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.10052158683538437, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.0207, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1014244481921196, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0205, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.15118759870529175, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0215, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.09245041012763977, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0176, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.13509149849414825, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0224, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.12747371196746826, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0185, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11445175856351852, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0208, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11949758231639862, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0193, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.12797918915748596, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0182, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0185, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.11956185847520828, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0189, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.11755053699016571, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0184, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.1273607611656189, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0205, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.11855774372816086, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0188, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.09031623601913452, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0183, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12208469212055206, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0199, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.1083436831831932, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0193, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12228160351514816, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0182, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.1007765457034111, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0195, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.11114434152841568, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0173, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.1096438392996788, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0199, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.124867744743824, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0219, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.10032925009727478, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.0194, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.13164062798023224, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0193, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.1376691609621048, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0187, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.11254525184631348, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0186, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.12835443019866943, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.15313562750816345, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0219, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.16989216208457947, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0172, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.11433102935552597, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.019, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09817333519458771, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0222, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.09500890970230103, + "learning_rate": 2.193514973491427e-05, + "loss": 0.0176, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.14314892888069153, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0199, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.11764178425073624, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0169, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.12384455651044846, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0193, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.11438048630952835, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0174, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.13921929895877838, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0182, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.16880372166633606, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0174, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.1554516702890396, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0194, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.15355393290519714, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0212, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.1179167702794075, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0195, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.13388752937316895, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0208, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.10997489839792252, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0181, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.11783649772405624, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0189, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13451549410820007, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0186, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.1339322179555893, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0179, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.09972941130399704, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0214, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.10101358592510223, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0176, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.09771978110074997, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0177, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.14234617352485657, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0184, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.10462982952594757, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0197, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.09772231429815292, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0226, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.12018870562314987, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0164, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.10980863869190216, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0199, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.11285397410392761, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0172, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.1067347601056099, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0183, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.1046559289097786, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.0198, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.11118492484092712, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0179, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.1001342311501503, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0166, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.1053142324090004, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0188, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.16253869235515594, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.018, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.10042452812194824, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0173, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.11815430223941803, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0181, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.07585619390010834, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0195, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.17831926047801971, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0183, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.09953836351633072, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0194, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.12129344046115875, + "learning_rate": 2.18065424687869e-05, + "loss": 0.0188, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.09982063621282578, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0196, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.0906965360045433, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0172, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.1256277859210968, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.019, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.10290617495775223, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0171, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.17007306218147278, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0178, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.0779242068529129, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0192, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.12276264280080795, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0233, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.13211409747600555, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0184, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11407056450843811, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0211, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.13331526517868042, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0193, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.1603303849697113, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0192, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.10634022206068039, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0174, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.1658177524805069, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0207, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.1432579606771469, + "learning_rate": 2.175492159090619e-05, + "loss": 0.02, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.11264102905988693, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0183, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.12988954782485962, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.0199, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.1687532663345337, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0175, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12261682003736496, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0148, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.13096439838409424, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0175, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.13871102035045624, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0198, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.12536168098449707, + "learning_rate": 2.172907339903944e-05, + "loss": 0.018, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.12913931906223297, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0194, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.1255621761083603, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0172, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.12288706004619598, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0206, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.113137386739254, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0201, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.11107201129198074, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.14609375596046448, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.02, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.15227735042572021, + "learning_rate": 2.170320016677318e-05, + "loss": 0.016, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.12586566805839539, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0194, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.13495711982250214, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0187, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.11436781287193298, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.018, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.12271122634410858, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0193, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.16221405565738678, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0195, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.08645932376384735, + "learning_rate": 2.168100325341058e-05, + "loss": 0.0157, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.11750698834657669, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0186, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.1191735714673996, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0178, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.10067173838615417, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0187, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.09761539846658707, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0169, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.13806267082691193, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0217, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.10618548095226288, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0165, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.15135695040225983, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0199, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10563432425260544, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0176, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.12757790088653564, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0183, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.11258719861507416, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.0162, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.15290242433547974, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0192, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.1350955218076706, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0188, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.09494426846504211, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0187, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12816165387630463, + "learning_rate": 2.162913952762737e-05, + "loss": 0.019, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.1237507089972496, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0187, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.12867869436740875, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0202, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11882596462965012, + "learning_rate": 2.161801300808089e-05, + "loss": 0.019, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.10291648656129837, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0187, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.09269602596759796, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.019, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.08865118771791458, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.0188, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10444393008947372, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0182, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.11461203545331955, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.0219, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.13728564977645874, + "learning_rate": 2.159574640654742e-05, + "loss": 0.0223, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.1394069343805313, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0172, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.15394151210784912, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0182, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.10422483086585999, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0175, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.10996434837579727, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0182, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.1026761382818222, + "learning_rate": 2.157717712784381e-05, + "loss": 0.018, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.1373908817768097, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0226, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.10155291855335236, + "learning_rate": 2.156974591793542e-05, + "loss": 0.0179, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.14904215931892395, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0192, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.1131916344165802, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0175, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.09735766798257828, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0182, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.13132493197917938, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0193, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.12934266030788422, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0184, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.1196565330028534, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0186, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.12375814467668533, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.0199, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.0892246812582016, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0178, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.0932748094201088, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0192, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11503782123327255, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0187, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.08250989019870758, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0183, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.101202592253685, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0183, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1226576566696167, + "learning_rate": 2.152139450890464e-05, + "loss": 0.022, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.1167614683508873, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0176, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.12396091967821121, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0186, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.13461633026599884, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.0169, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.10792742669582367, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0189, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11649876832962036, + "learning_rate": 2.150277549265773e-05, + "loss": 0.02, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.1314917951822281, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0157, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.14143578708171844, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0197, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10042720288038254, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0177, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.12019779533147812, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0164, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.14096254110336304, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0223, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.10627058893442154, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0171, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.11092417687177658, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.019, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.12211407721042633, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0199, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11551571637392044, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0197, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.10474767535924911, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.018, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.12364303320646286, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0178, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13243450224399567, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0197, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10594736039638519, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0179, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.11143538355827332, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0196, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.10518378764390945, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0197, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12268701940774918, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0176, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.09408178925514221, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0205, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.12529458105564117, + "learning_rate": 2.143564506261267e-05, + "loss": 0.0184, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.15819092094898224, + "learning_rate": 2.143191093594534e-05, + "loss": 0.021, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.15818199515342712, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0182, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.13374921679496765, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0177, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.10046645253896713, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0197, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.11422070860862732, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0185, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.1138019785284996, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0182, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.09617039561271667, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0178, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.1452113538980484, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0222, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.10409227013587952, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0196, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.11277221143245697, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0175, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.12325838208198547, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0201, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12721778452396393, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0174, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.14016005396842957, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.017, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.11149191111326218, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0168, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.13445520401000977, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0194, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10086208581924438, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0197, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.115168996155262, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0181, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.11483300477266312, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.0183, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.1196938082575798, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0172, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.12219196557998657, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0177, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.11318044364452362, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.0204, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.09219531714916229, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0181, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.09911682456731796, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0192, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.11822745203971863, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0192, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.11101230978965759, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0165, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.09978491812944412, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0203, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.09874484688043594, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0199, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.08682837337255478, + "learning_rate": 2.133090580749213e-05, + "loss": 0.0177, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.14352355897426605, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0189, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.11133426427841187, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0199, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.10667043924331665, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0183, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.14207594096660614, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0218, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.12165568023920059, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0167, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.09311263263225555, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0183, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10423581302165985, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0151, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.12640471756458282, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0214, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.11349574476480484, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0171, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.14359329640865326, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0182, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.10977838933467865, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0181, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.12443531304597855, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0168, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.12441008538007736, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0186, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.14103129506111145, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0197, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10460726171731949, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0178, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.12526848912239075, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0188, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.13363681733608246, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0183, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.13106584548950195, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0181, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.1353096216917038, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0202, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.11889169365167618, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0198, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.12628430128097534, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0178, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.13260240852832794, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0182, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12357809394598007, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0175, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.13913387060165405, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0189, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.12303008884191513, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0187, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.10231908410787582, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.017, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.20065239071846008, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0172, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.11311347782611847, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.019, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.131951242685318, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0182, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.15945149958133698, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0202, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09272884577512741, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0189, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.0937163308262825, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0208, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.10963460803031921, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0197, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.1618727743625641, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0195, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.0919048935174942, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0177, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.12060586363077164, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0209, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.13121892511844635, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0181, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.11795026063919067, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0186, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.13060589134693146, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0226, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.08142203092575073, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0181, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.12634120881557465, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0178, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.1448773890733719, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0179, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1230989471077919, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0188, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.11944684386253357, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0212, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.1445692479610443, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0184, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.0831015482544899, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0171, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.1456967294216156, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0189, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.11207464337348938, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0185, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.16137553751468658, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0182, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.09820980578660965, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0192, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.11823676526546478, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0183, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.12873874604701996, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0181, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.11243417114019394, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0188, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.11120739579200745, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0222, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.10704528540372849, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0178, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.12063440680503845, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0174, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.13275623321533203, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0171, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.10400331765413284, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0201, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.10485411435365677, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0187, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.12517869472503662, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0187, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.11703156679868698, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.0178, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.11982452869415283, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0186, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11858765780925751, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0182, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.10404367744922638, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0173, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.09697552025318146, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0188, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.12046534568071365, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0173, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.0896986722946167, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0188, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.10993069410324097, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0213, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.1200791671872139, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0179, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.14664512872695923, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.0202, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.09285914152860641, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0207, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.1080656424164772, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0188, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.140005961060524, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0188, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.11084222793579102, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.02, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.0913061872124672, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0194, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.09459970891475677, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0182, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.09709502756595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0195, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.11101861298084259, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0184, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.1152782216668129, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0178, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.11495838314294815, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0178, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.14250393211841583, + "learning_rate": 2.10258099183954e-05, + "loss": 0.019, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.09898682683706284, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0173, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.13904057443141937, + "learning_rate": 2.101823807728166e-05, + "loss": 0.018, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.12013959139585495, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0163, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.0925723984837532, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0187, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.1284961700439453, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0158, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.12408998608589172, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0206, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1095876470208168, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0172, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.10166801512241364, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0183, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13473083078861237, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0155, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.10492225736379623, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0188, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.10769215226173401, + "learning_rate": 2.098414220174082e-05, + "loss": 0.021, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.09331739693880081, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0176, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.09247364103794098, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0163, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.20951682329177856, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0203, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.16750548779964447, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0196, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.10933154821395874, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0169, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10767697542905807, + "learning_rate": 2.09613911545763e-05, + "loss": 0.0184, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.11565674096345901, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0208, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.11934554576873779, + "learning_rate": 2.095380384728948e-05, + "loss": 0.018, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.12786556780338287, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0169, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.11752485483884811, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0194, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.11586920917034149, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0179, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.08849998563528061, + "learning_rate": 2.0938623809196e-05, + "loss": 0.018, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.12535157799720764, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0176, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.10429015755653381, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0188, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.09504082053899765, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0186, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.12917572259902954, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0182, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.11084037274122238, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0163, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.12090355157852173, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0196, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.11049258708953857, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0194, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.0831870436668396, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0188, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11006250977516174, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.018, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.12107989192008972, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0178, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.10361241549253464, + "learning_rate": 2.089684155251222e-05, + "loss": 0.017, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.13428105413913727, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0198, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.13375632464885712, + "learning_rate": 2.088923894557553e-05, + "loss": 0.018, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.14561115205287933, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0181, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.09850531816482544, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0176, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.11646582931280136, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0162, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.14659325778484344, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0168, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.15971294045448303, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0198, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.11780509352684021, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0156, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11453750729560852, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0179, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.09094218909740448, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0172, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.13252301514148712, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0209, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.10068004578351974, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0159, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.10302501171827316, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0172, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.10893740504980087, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.019, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.12446500360965729, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0177, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1345142424106598, + "learning_rate": 2.083597073454921e-05, + "loss": 0.021, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.12400489300489426, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0194, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.15042100846767426, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0162, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.1559734046459198, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0189, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.10689649730920792, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0163, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.15904614329338074, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0199, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.11539188027381897, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0208, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.14550678431987762, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0191, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.10212639719247818, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0177, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.10318224877119064, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0176, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0196, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.10269734263420105, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0179, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.11070621758699417, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0196, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13299307227134705, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.02, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11430692672729492, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0188, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10718249529600143, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.0204, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.11855681240558624, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.019, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.10754020512104034, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0157, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11011523753404617, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0172, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.11931152641773224, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.0171, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.12425696104764938, + "learning_rate": 2.075972279799841e-05, + "loss": 0.0195, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.10292454063892365, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0177, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.11290059238672256, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0165, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.1453622728586197, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0207, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.12152253836393356, + "learning_rate": 2.07444521407078e-05, + "loss": 0.017, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.134286567568779, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0198, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.11180262267589569, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0196, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.12045837193727493, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.018, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.18509334325790405, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0209, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.17120368778705597, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0197, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.11996874213218689, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0204, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.12391912192106247, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0201, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.1248779147863388, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.016, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.12873846292495728, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0184, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.11072682589292526, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.0182, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.10367089509963989, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0179, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.12854290008544922, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0184, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.16348853707313538, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0169, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.1345866173505783, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0187, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.1139485239982605, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0178, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.1656051129102707, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0177, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.1133129820227623, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0154, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.0865294560790062, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0163, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.1368715763092041, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0218, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.11601361632347107, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0185, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.1401238590478897, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0177, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.10183753073215485, + "learning_rate": 2.066033925513678e-05, + "loss": 0.0192, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.10317160934209824, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0153, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.12317138910293579, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.021, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.10953490436077118, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0191, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.10530596971511841, + "learning_rate": 2.064502355956822e-05, + "loss": 0.0179, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.09054628014564514, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0189, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.12472135573625565, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0239, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.1551167368888855, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0189, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.08141466975212097, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0174, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.13025878369808197, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0185, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.13056428730487823, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0183, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.1187913790345192, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0182, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.11957527697086334, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0173, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.09339714050292969, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0182, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.1584015190601349, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.02, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.13754191994667053, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0193, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.09880727529525757, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0192, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.15097497403621674, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0172, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12028415501117706, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0182, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.10583297163248062, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0183, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.09343970566987991, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0187, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.1436399519443512, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0175, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.10790709406137466, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0184, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.14520230889320374, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0187, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.10826292634010315, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0188, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.1122521162033081, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0197, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.11445961147546768, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.015, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.1473265290260315, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0193, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.1572754830121994, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0186, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.1085842102766037, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0169, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.11982112377882004, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0163, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.12014015763998032, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.0217, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.11440391838550568, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0175, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.1276320368051529, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0203, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12434501200914383, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0186, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.11981593072414398, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0138, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.08389589190483093, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0162, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.11693917959928513, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.0199, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.1033058762550354, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0182, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.17878879606723785, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0181, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.12847506999969482, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0186, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.10514750331640244, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0166, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.1037413701415062, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0182, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.10402073711156845, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0165, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.10330315679311752, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0164, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.14082680642604828, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0205, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.1187078058719635, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0178, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.10590347647666931, + "learning_rate": 2.047995045386578e-05, + "loss": 0.0192, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.11178472638130188, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.019, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.11713241040706635, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0169, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.1520795077085495, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0173, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.11906655877828598, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0163, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.10854434967041016, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0181, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.16785599291324615, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.0212, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10305774211883545, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0179, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.09346644580364227, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0162, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11048682034015656, + "learning_rate": 2.044530211206255e-05, + "loss": 0.0158, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.14503270387649536, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0185, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.12631897628307343, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0182, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.14716260135173798, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0205, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.1364247053861618, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0187, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.09775770455598831, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0179, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.17999885976314545, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0194, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.1381615400314331, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0192, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.10749474912881851, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0189, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.15220865607261658, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.12083570659160614, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0175, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.11410468816757202, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0174, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.11285928636789322, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.0191, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.09808669239282608, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0173, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.07822877913713455, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0179, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.1032382920384407, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0177, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12453432381153107, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0183, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.1397278904914856, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0175, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.1701551228761673, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0176, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.13178907334804535, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0158, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.16403238475322723, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0182, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.13050618767738342, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.019, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09723208099603653, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.016, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.11937404423952103, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0194, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.10595990717411041, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0179, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.10598801076412201, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0172, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.12562280893325806, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0192, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.1062425896525383, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0167, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.19852802157402039, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0186, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10752324014902115, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0171, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.11313100904226303, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0191, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.09881386160850525, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0189, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.11926184594631195, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0176, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.11304130405187607, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0171, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.159687340259552, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0209, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.12324368953704834, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0184, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.10718217492103577, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0178, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12370229512453079, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0181, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.11843078583478928, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.017, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.11866020411252975, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0173, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.10999849438667297, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0217, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.10843008756637573, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.0163, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.13494807481765747, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0202, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.13634233176708221, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.0203, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.09910089522600174, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0154, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.1105995923280716, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0172, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.11397486180067062, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0171, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10141155868768692, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0187, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.09431210905313492, + "learning_rate": 2.025995044916193e-05, + "loss": 0.0163, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.09627058357000351, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0188, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.10498029738664627, + "learning_rate": 2.025220726516947e-05, + "loss": 0.0197, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.12692295014858246, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0163, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.11992908269166946, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0188, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.1448463350534439, + "learning_rate": 2.024058949838742e-05, + "loss": 0.0182, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.10764076560735703, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0191, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.1176103726029396, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0202, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.10393970459699631, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0174, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.12357908487319946, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0188, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11184806376695633, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0171, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09287009388208389, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0164, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.13145606219768524, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0175, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.108634814620018, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0192, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.1643940508365631, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.02, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.1642858386039734, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0195, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.11515326052904129, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0173, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.11438507586717606, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0187, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.1556805521249771, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0193, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.1081680878996849, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0176, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.11440131068229675, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0198, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.13953827321529388, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0163, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.12443852424621582, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0173, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.12884436547756195, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.0184, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.09170050173997879, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0158, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.08685769885778427, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0143, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.12880417704582214, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0172, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.09957600384950638, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0189, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.11359020322561264, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0191, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.10620398819446564, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0183, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.11106064170598984, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0175, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.14327868819236755, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0156, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.09743347764015198, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0165, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.11041118204593658, + "learning_rate": 2.013198534613597e-05, + "loss": 0.019, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.13757234811782837, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.017, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.10167638957500458, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0172, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.1386769413948059, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0157, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.13493874669075012, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0184, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.1238163560628891, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0178, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12875531613826752, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0176, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11199075728654861, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0179, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12402849644422531, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0197, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.12012362480163574, + "learning_rate": 2.009701166604228e-05, + "loss": 0.0203, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.1319619119167328, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0176, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.11875088512897491, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.017, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.14083963632583618, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0192, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1157609224319458, + "learning_rate": 2.008145773123833e-05, + "loss": 0.017, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1447380781173706, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0192, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.08668482303619385, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.0182, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.11417029052972794, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0162, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.12431302666664124, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0176, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.09466125816106796, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0174, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.13191643357276917, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0162, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.16682866215705872, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0167, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.09262735396623611, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0174, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10246162116527557, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0157, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.14533615112304688, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0193, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10985267907381058, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0185, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.1364467293024063, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0174, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.12792983651161194, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.019, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.13055621087551117, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0175, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.1272381991147995, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0191, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.1523309350013733, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.0192, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.12438428401947021, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0183, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.13417088985443115, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0192, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.10955093055963516, + "learning_rate": 2.000749254391884e-05, + "loss": 0.019, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.1140032708644867, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.0155, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.13144028186798096, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.0177, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.10048601031303406, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0169, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.1272922307252884, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0167, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.10301118344068527, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0166, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.10341407358646393, + "learning_rate": 1.998410649877165e-05, + "loss": 0.018, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.08479570597410202, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0193, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.10605745017528534, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0169, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.09806226938962936, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0175, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.12049946188926697, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0187, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.13633446395397186, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0167, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.12669390439987183, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.02, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.1395879089832306, + "learning_rate": 1.99568055617237e-05, + "loss": 0.02, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.1145915761590004, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.019, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12087064981460571, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0194, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.11252355575561523, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0156, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.10857690125703812, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0197, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.17054887115955353, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0172, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.12123709172010422, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0184, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.08856352418661118, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0178, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.09841718524694443, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0198, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.1236206591129303, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0171, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11047739535570145, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0198, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.11695606261491776, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0162, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.12013910710811615, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0171, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.08011548966169357, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0176, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.13469302654266357, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0177, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.12169095873832703, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.0192, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.1165102869272232, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0165, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.16436143219470978, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0164, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.11628631502389908, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0174, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.13279633224010468, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0164, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11530247330665588, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0199, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13026569783687592, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0193, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.12264525890350342, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0178, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.09320709109306335, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0157, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.11837475001811981, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.016, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.12293318659067154, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.0182, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.09163888543844223, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0197, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.1171371266245842, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0162, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.1096082329750061, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0198, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09364283084869385, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0164, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.08231621980667114, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0169, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.1273244321346283, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.0167, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.15531282126903534, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0194, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.13366153836250305, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0186, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.1072721853852272, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0173, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.11176268011331558, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0182, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.10074266046285629, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0167, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.11638782918453217, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0194, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.1109209880232811, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0197, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.1511884480714798, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0202, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.09859053790569305, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0185, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.12002220004796982, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0177, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.11930137127637863, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0159, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.10564088076353073, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0176, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.10428094118833542, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0197, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.1335987150669098, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0188, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.12865577638149261, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0179, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.14246998727321625, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0181, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.1250566691160202, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0188, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.09627831727266312, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0153, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.09008506685495377, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0185, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.12598916888237, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0169, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.11335116624832153, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0165, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.10588988661766052, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0171, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.10731689631938934, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.0174, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.11184538155794144, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0154, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.10870286822319031, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0163, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.11692065745592117, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0178, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.10036200284957886, + "learning_rate": 1.972989772246684e-05, + "loss": 0.016, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.12077154219150543, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0178, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.1272457093000412, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0188, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.08074570447206497, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0159, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.11381031572818756, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.0155, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.09836770594120026, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0161, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.09236133843660355, + "learning_rate": 1.970635454034925e-05, + "loss": 0.0183, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.1167517825961113, + "learning_rate": 1.970242942396647e-05, + "loss": 0.018, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.10953414440155029, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.015, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.12463351339101791, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0192, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.1285952925682068, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0172, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.10794857144355774, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0167, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.11815376579761505, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0176, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.10251157730817795, + "learning_rate": 1.967887123661664e-05, + "loss": 0.018, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.16941046714782715, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0182, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.14373992383480072, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.0177, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.11837214231491089, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0169, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.10599380731582642, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0167, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.11006144434213638, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0207, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.11699084937572479, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0164, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.10131341218948364, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0177, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.09975451976060867, + "learning_rate": 1.964744043284577e-05, + "loss": 0.017, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.08056963980197906, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0171, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.16518834233283997, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0186, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.10148414224386215, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0181, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.14867499470710754, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0183, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.16952896118164062, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0199, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13692231476306915, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0198, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.15133927762508392, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0184, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11434897035360336, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0189, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.0917418971657753, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0167, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12218283116817474, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0168, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.12193279713392258, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0173, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.09389539062976837, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0182, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.10455376654863358, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0169, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.11151516437530518, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0188, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.14512453973293304, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0194, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.10494859516620636, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0169, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.15839508175849915, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0177, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.12615326046943665, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0172, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.1135321632027626, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0166, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.12872013449668884, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0173, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.18429434299468994, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0165, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.13249747455120087, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0183, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.10603360831737518, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0145, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.10620120912790298, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0155, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.10986930131912231, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0163, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09298109263181686, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0179, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09945658594369888, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0178, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.08814677596092224, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0186, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.09042713046073914, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0173, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.11592321842908859, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0191, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.1026391014456749, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.018, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.09040834754705429, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0164, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.11391181498765945, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0177, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11113253980875015, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.0166, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.10178712755441666, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0167, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.13696520030498505, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0171, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.10290519148111343, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0164, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13472187519073486, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0173, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.08172356337308884, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0187, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.12297531962394714, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0166, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.13925355672836304, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0193, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.15394000709056854, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0169, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.09050899744033813, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.019, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.11854210495948792, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0181, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.10207716375589371, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0154, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.10006722062826157, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0161, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.10565274208784103, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0183, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.11656428128480911, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0166, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.12020296603441238, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0181, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.0964578241109848, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0174, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.0888240784406662, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0158, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.09271078556776047, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0187, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.10407985746860504, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0152, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1329568326473236, + "learning_rate": 1.94346993525268e-05, + "loss": 0.0166, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.13284213840961456, + "learning_rate": 1.943075029574398e-05, + "loss": 0.018, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.14158464968204498, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0194, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.11495806276798248, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0157, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.12886112928390503, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0174, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08954162895679474, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0158, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.09367296099662781, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0171, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.11188989877700806, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0162, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.10758768022060394, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0154, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.11783428490161896, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0182, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.18594273924827576, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0175, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.08010479807853699, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0167, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13949649035930634, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0175, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.10709669440984726, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0179, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.1675017923116684, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0176, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12764538824558258, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0165, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11073391884565353, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0169, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.10426445305347443, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0171, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.09000374376773834, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0161, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.10287991911172867, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0171, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.12322332710027695, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0178, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.11188322305679321, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0166, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.14056499302387238, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0196, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.08969348669052124, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0169, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.10100627690553665, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0159, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.1181580200791359, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0185, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.11281560361385345, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0171, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.11793895065784454, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0187, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.14534896612167358, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0177, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.10177207738161087, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0182, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.1335727721452713, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.017, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.14608687162399292, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0197, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.09763725847005844, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0184, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.09486788511276245, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0177, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.12235541641712189, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.018, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.13838450610637665, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0174, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.09837213158607483, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.0174, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.11713139712810516, + "learning_rate": 1.928836248385875e-05, + "loss": 0.016, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.13015268743038177, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0187, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.13498486578464508, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0181, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.12189928442239761, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0156, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.14399787783622742, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0185, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.10662715882062912, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0164, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.1145375519990921, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0185, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.11677944660186768, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0153, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.10242128372192383, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0168, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.1290024369955063, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0189, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.11820637434720993, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0167, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.12113048881292343, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0187, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.12122460454702377, + "learning_rate": 1.92408053936279e-05, + "loss": 0.0175, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.09413880109786987, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0149, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.11389796435832977, + "learning_rate": 1.923287468901755e-05, + "loss": 0.0185, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13923633098602295, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.0169, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.15947312116622925, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0188, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.14314432442188263, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0182, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.1426798403263092, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0179, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.12179753184318542, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.02, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11758018285036087, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0197, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11349253356456757, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0171, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.13724955916404724, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0181, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.1260363608598709, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0172, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.09541817009449005, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0181, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.0922083631157875, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0177, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.08051180839538574, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0167, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.11803245544433594, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0177, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10018756985664368, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.10512874275445938, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0164, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.17478078603744507, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0169, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.08864858746528625, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0165, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11313372850418091, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0172, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.09432164579629898, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0163, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.09435626119375229, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0188, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.10041314363479614, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0157, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.07982632517814636, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0176, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.09699293971061707, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0177, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.11059583723545074, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.017, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.11384644359350204, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0163, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.11473467200994492, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0165, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.11018316447734833, + "learning_rate": 1.912568530871305e-05, + "loss": 0.0179, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.10781396180391312, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0199, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.11702359467744827, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0165, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.12909658253192902, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0162, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.09839031100273132, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0152, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.1357898861169815, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0198, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.11221814155578613, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.0178, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.11431093513965607, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0176, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.11436092853546143, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0187, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.09838918596506119, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0148, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.11489785462617874, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0168, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.10876795649528503, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0169, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.1032189428806305, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0152, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.10013218969106674, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.11554815620183945, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0182, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.2529490292072296, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0198, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.09215852618217468, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0168, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.12875677645206451, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0185, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.1232401430606842, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0179, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.10334797203540802, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0171, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10524798929691315, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0162, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.11014945805072784, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0176, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.1074177473783493, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0162, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.14043818414211273, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0191, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.10011711716651917, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0178, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0172, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.12526027858257294, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0137, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.15774008631706238, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0153, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.1049765795469284, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0155, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09307108074426651, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0186, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1369069367647171, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0158, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.12320058792829514, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0181, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.10885081440210342, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0168, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.08989177644252777, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0153, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.09219451993703842, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0172, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.11661440879106522, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0192, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.0971342995762825, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0178, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.10453730076551437, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0165, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.17652514576911926, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0196, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10733854025602341, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0154, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.0989762544631958, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0181, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.12107723951339722, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.1385297179222107, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0191, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.12209209054708481, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0217, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.09332361072301865, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0158, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.12448447197675705, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0164, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.12061425298452377, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0169, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.12132656574249268, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0183, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.14007116854190826, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0167, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.11448334902524948, + "learning_rate": 1.893057716200171e-05, + "loss": 0.015, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.12702956795692444, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0152, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.09236849099397659, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0177, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.09818986803293228, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.0164, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12131274491548538, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0171, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10609138011932373, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.0147, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.1431477814912796, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0173, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.0844314768910408, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0162, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.09800038486719131, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0157, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.11529140919446945, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0172, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.086520716547966, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0154, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.12068255245685577, + "learning_rate": 1.888667760339568e-05, + "loss": 0.015, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.12278824299573898, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0174, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.10242115706205368, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0173, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.11696609109640121, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0187, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.11133866757154465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0151, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.1319858729839325, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0165, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.08355733752250671, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0166, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.10663259029388428, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0177, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.09787473827600479, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0154, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.12111647427082062, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0171, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.10036861151456833, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.15351398289203644, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.019, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.11561891436576843, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0174, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.09197179228067398, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0174, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.0961032509803772, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.019, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.09264974296092987, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0158, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.11056753247976303, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0159, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.10464371740818024, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0165, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.13888190686702728, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0185, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.09801805019378662, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0189, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.15648017823696136, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0167, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1129915863275528, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0166, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.10456246137619019, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0171, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.10873697698116302, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.0168, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.17470459640026093, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0169, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.15504027903079987, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.10821744054555893, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0152, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.14031334221363068, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0181, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.128667414188385, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0183, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.09095419198274612, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0151, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.09303424507379532, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0167, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.10706666111946106, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0174, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.13011431694030762, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0177, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.13821059465408325, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0154, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.10680826753377914, + "learning_rate": 1.875076403503638e-05, + "loss": 0.017, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.12847846746444702, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0146, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.1472046673297882, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0184, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10622651875019073, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0153, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.129075825214386, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0189, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.12048091739416122, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.017, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.10954522341489792, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0175, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.10811172425746918, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0173, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.13378627598285675, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0193, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.13185591995716095, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0195, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.1013568863272667, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0182, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.07558644562959671, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0173, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.10668148100376129, + "learning_rate": 1.870271530813327e-05, + "loss": 0.018, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.134077250957489, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0157, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.14325739443302155, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0221, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.1389995515346527, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0178, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.13043907284736633, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0169, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.12082460522651672, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0174, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11521385610103607, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0165, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.11550573259592056, + "learning_rate": 1.867466815081593e-05, + "loss": 0.017, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.11315014213323593, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0173, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.1067388728260994, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0161, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.0941673144698143, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0158, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.09685055166482925, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0161, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.11744765937328339, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0164, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.10552993416786194, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0154, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.12654545903205872, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0184, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.11333387345075607, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0186, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.09856864809989929, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.0187, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.0933985561132431, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0171, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.096091628074646, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0194, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.08463368564844131, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0165, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.10648614913225174, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.017, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.10691184550523758, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0165, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1005169153213501, + "learning_rate": 1.861452118736741e-05, + "loss": 0.0174, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.1037023589015007, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0172, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.1299487054347992, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.015, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14639562368392944, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0167, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.12270855903625488, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0173, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.09706515073776245, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0169, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.11888926476240158, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.0152, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.11289773136377335, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0175, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13422279059886932, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0168, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.10992032289505005, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0176, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.10719195753335953, + "learning_rate": 1.85743888359202e-05, + "loss": 0.017, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.10348568856716156, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0166, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.10436776280403137, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0172, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.09779159724712372, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.018, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.10799746960401535, + "learning_rate": 1.855832827763672e-05, + "loss": 0.016, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.11972185224294662, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0195, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.10282087326049805, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0169, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.11840159446001053, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0183, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.13861460983753204, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0205, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10774466395378113, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0163, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.14446309208869934, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0168, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.1313662976026535, + "learning_rate": 1.853021190846946e-05, + "loss": 0.0161, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.0857798308134079, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0159, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.11933837085962296, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0178, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11762946099042892, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0154, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.11128528416156769, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0198, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.07360108196735382, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0148, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.10680225491523743, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0174, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.11387799680233002, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0188, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.2774786651134491, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0165, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.12314122170209885, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.017, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.14403507113456726, + "learning_rate": 1.849002290456207e-05, + "loss": 0.0175, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.10811382532119751, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0173, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.1399172991514206, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0154, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.10577643662691116, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0177, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.10794928669929504, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0155, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.0844842791557312, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0149, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.09744124114513397, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0166, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.14749610424041748, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0179, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.09801949560642242, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0177, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.07795082777738571, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0174, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.11197780072689056, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.0172, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.11819472163915634, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.08461402356624603, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0171, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.11102861911058426, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0156, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.11057187616825104, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0189, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.11031147837638855, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0155, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.08435497432947159, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0187, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.12450679391622543, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.017, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.09348208457231522, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0171, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.10584745556116104, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0161, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.14194349944591522, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.0177, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.1025288999080658, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0153, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.07800278812646866, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0157, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.1253945231437683, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0176, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.11724173277616501, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.016, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.13043718039989471, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0179, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.08728159964084625, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0177, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.1419975906610489, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0185, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.12774096429347992, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0187, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.10763564705848694, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.0163, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.12671130895614624, + "learning_rate": 1.836929808825424e-05, + "loss": 0.0177, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.10261370241641998, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0145, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12106029689311981, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.0157, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.12428469955921173, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0194, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.08797722309827805, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0175, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.14663904905319214, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.018, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.12288253754377365, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0176, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.11406450718641281, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0164, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10584931075572968, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.017, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1212797686457634, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0176, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.10806597024202347, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0159, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.10098059475421906, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0186, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.14030125737190247, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0204, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.09907040745019913, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0158, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1299905776977539, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0198, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.134843647480011, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0157, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.1048227846622467, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0154, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.12262892723083496, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0193, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.12806116044521332, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0167, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.12875764071941376, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.0149, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.09743807464838028, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0149, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10791302472352982, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0155, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10845673829317093, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0187, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.1366390436887741, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0161, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.101824089884758, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0174, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.12938271462917328, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0159, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.13503864407539368, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0196, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10241002589464188, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0153, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.10051950812339783, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.0159, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11316224932670593, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0166, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.10802372545003891, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0167, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.1365254521369934, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0176, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.11756277829408646, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.016, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13363559544086456, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0167, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.13138234615325928, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0194, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.1519714891910553, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0147, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.12152288109064102, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.017, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.0976492166519165, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0174, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.09356466680765152, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0158, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.09382101893424988, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0175, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.12188217788934708, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0163, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.09245157241821289, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0183, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.1661074161529541, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0202, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.154639333486557, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0143, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.117487333714962, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.017, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.09468945115804672, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0171, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.09632660448551178, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0147, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.12390028685331345, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0177, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.11924377828836441, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0154, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.13191300630569458, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0154, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.07403766363859177, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0171, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.10911701619625092, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0159, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.11888141930103302, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0189, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.09766717255115509, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0182, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.08285287767648697, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0145, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.13635197281837463, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0166, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.11627981066703796, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.10899969190359116, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0174, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.10002870857715607, + "learning_rate": 1.813525107780979e-05, + "loss": 0.016, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09963744133710861, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0163, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.2032729536294937, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0188, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.10394410789012909, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0178, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.1004093736410141, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0181, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.08776404708623886, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.0188, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.10926580429077148, + "learning_rate": 1.811099258182467e-05, + "loss": 0.018, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.09496597945690155, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0153, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.12430521845817566, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0152, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.11681065708398819, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0156, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.10450825840234756, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0162, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10932883620262146, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0163, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.1358717978000641, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0152, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.11744217574596405, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0188, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.14966122806072235, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0179, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.113619364798069, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.08801224827766418, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0168, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.09556800872087479, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0171, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.11116721481084824, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0171, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.10650353878736496, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0158, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.09222083538770676, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0154, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.10958139598369598, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.0189, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11020398885011673, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0166, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.08026755601167679, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0159, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.09629099816083908, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0162, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11310935020446777, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0192, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.1281706690788269, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0191, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12402708828449249, + "learning_rate": 1.802602130181036e-05, + "loss": 0.0159, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.09760920703411102, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0151, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.12342818826436996, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.018, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09423290938138962, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0163, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.10432913154363632, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0172, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.09628765285015106, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0169, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.11590366810560226, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0167, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10148236900568008, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0155, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.10599179565906525, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0155, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.11854606866836548, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.10626410692930222, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0185, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.12172172963619232, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0179, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.10842256247997284, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0173, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.11806859076023102, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0163, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11489225924015045, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0178, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.11859364807605743, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0164, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.18860413134098053, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.0168, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.13375963270664215, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0168, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.13163933157920837, + "learning_rate": 1.795310800633073e-05, + "loss": 0.0151, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.11135797202587128, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0175, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.12392686307430267, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0161, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.0995136946439743, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0165, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.10128366202116013, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0161, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.13080070912837982, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0169, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.10262267291545868, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0176, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.10983351618051529, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0163, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.09834928810596466, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0158, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.11170317232608795, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0166, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.14308896660804749, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.09110505878925323, + "learning_rate": 1.790851399590954e-05, + "loss": 0.017, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.12270113825798035, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0165, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.1422351896762848, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0177, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1294422596693039, + "learning_rate": 1.789634733729274e-05, + "loss": 0.017, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.1157895103096962, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0192, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.12473416328430176, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0156, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.1765025109052658, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0161, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11862069368362427, + "learning_rate": 1.78801220478228e-05, + "loss": 0.017, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.10549142211675644, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0165, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.09818614274263382, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0171, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.13607585430145264, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0168, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.09340739995241165, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0158, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.0805685818195343, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.0207, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.11172603815793991, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0164, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.08757263422012329, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0179, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.09590335935354233, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0199, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.1186867505311966, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.0146, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.09467568248510361, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0175, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11578786373138428, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0179, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09604159742593765, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0173, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.11337848007678986, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.0156, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.11909373849630356, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0188, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.08804072439670563, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0175, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.13454975187778473, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.0184, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.12201394885778427, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0148, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.11946159601211548, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0153, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12322071939706802, + "learning_rate": 1.78030044678245e-05, + "loss": 0.0181, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.10444212704896927, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0159, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.10723353922367096, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0158, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.10750993341207504, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0167, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.1207774356007576, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0207, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.12730205059051514, + "learning_rate": 1.778269750931628e-05, + "loss": 0.017, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.08926276862621307, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0151, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.10322589427232742, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0174, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10034044831991196, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.0184, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.12131599336862564, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0183, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.10715801268815994, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0156, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.14093570411205292, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0169, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.124595507979393, + "learning_rate": 1.775425890028875e-05, + "loss": 0.017, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.13153409957885742, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0189, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.1348494291305542, + "learning_rate": 1.77461316962853e-05, + "loss": 0.019, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.10926292091608047, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0161, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.11094044893980026, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0155, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.1341349333524704, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0184, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.15732111036777496, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0203, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.07597270607948303, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.0184, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.11097310483455658, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.0176, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.12087065726518631, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0164, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12998445332050323, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0191, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.13928557932376862, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0157, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.10487896203994751, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0172, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.13437892496585846, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0182, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.20842674374580383, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0189, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.09813268482685089, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12833304703235626, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0168, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.1019122451543808, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0159, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.07276969403028488, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.016, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09712188690900803, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0169, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.10403666645288467, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0156, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.11877460777759552, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0152, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.10731421411037445, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0188, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.12387207895517349, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0173, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.10561594367027283, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0179, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.2772648334503174, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0168, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.097810760140419, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0142, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.1168239563703537, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0171, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.14195922017097473, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0162, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11385945230722427, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0187, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.1124262586236, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.13074515759944916, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0151, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.12591321766376495, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0177, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.09103676676750183, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0169, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.12457307428121567, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0162, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.11226223409175873, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0156, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.14778819680213928, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0187, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.10433483868837357, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0187, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.08798553794622421, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0163, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.09729046374559402, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0149, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10188018530607224, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0154, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.12559989094734192, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.018, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12567821145057678, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0168, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.12211210280656815, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0166, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.11825426667928696, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0168, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.11676980555057526, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0175, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.11447525024414062, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0162, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.08958559483289719, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0161, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.11528483778238297, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0162, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.08296159654855728, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.014, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.12306056916713715, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0158, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.13315193355083466, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.0164, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.09533339738845825, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0166, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.20049314200878143, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.017, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.14355438947677612, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0187, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.11443490535020828, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0146, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.10336155444383621, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0164, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11926902830600739, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0173, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.09699229151010513, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0176, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.11406200379133224, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0164, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.10817506164312363, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0165, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.11398246139287949, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.0163, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.09427376091480255, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0166, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.10372194647789001, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0148, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.09003781527280807, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0172, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.09807098656892776, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.11895686388015747, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0163, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10839524865150452, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0173, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.20792271196842194, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0183, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.07746267318725586, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0147, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.11470094323158264, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0177, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.1377800703048706, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.1247936338186264, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.016, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.1405080407857895, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0164, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.1318410336971283, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0175, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.14254692196846008, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0159, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.12803269922733307, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0146, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11180103570222855, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0169, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.16513720154762268, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0162, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.13401633501052856, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.0182, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.13510771095752716, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0175, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.12391090393066406, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0174, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1019999086856842, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0157, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.11906205862760544, + "learning_rate": 1.7416306237948e-05, + "loss": 0.0179, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11846558749675751, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.0154, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.12292807549238205, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0163, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10195571929216385, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0186, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.1032365933060646, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0163, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1241045668721199, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0173, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.10956405848264694, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.0163, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.12094716727733612, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.018, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.1353612244129181, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0161, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.0935104638338089, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0164, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.10712502896785736, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.0177, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.12715180218219757, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0148, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.09601753205060959, + "learning_rate": 1.736733684019633e-05, + "loss": 0.0167, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.11189524084329605, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0174, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.11025495082139969, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.016, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.1203494742512703, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0192, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.12150877714157104, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0143, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.11957377195358276, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0146, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.16482087969779968, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0181, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.10647732019424438, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0178, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.13345423340797424, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0155, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.13268540799617767, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0178, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.12209953367710114, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0161, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.11951886862516403, + "learning_rate": 1.73224254699117e-05, + "loss": 0.0213, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.12044327706098557, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0195, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.10611505806446075, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0169, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.12347753345966339, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0145, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13930706679821014, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0185, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.12702012062072754, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0167, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.1193593367934227, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0169, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.16778719425201416, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0179, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11699119210243225, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0159, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11571445316076279, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.08481728285551071, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0159, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.10007977485656738, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0155, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.12851299345493317, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0167, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.13929468393325806, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0154, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.11237664520740509, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0143, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.11654951423406601, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0179, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.11065302044153214, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0168, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.10024461150169373, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0173, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.1334248036146164, + "learning_rate": 1.724888829234145e-05, + "loss": 0.016, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.1115458682179451, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0159, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.15415695309638977, + "learning_rate": 1.72407140459686e-05, + "loss": 0.0166, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10965333878993988, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0159, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.10152038931846619, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0152, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.10030021518468857, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0161, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.1526176780462265, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0134, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13154515624046326, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.017, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.12260743975639343, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0152, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.07395336031913757, + "learning_rate": 1.721209883968656e-05, + "loss": 0.0155, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.1107548177242279, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0178, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.08634423464536667, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0148, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.10544566810131073, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0151, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.13896141946315765, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0171, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10351476818323135, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0164, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.11912661790847778, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0163, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.12101101130247116, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0185, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.09519081562757492, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0137, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.13252407312393188, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0153, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.08964477479457855, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0148, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.11772924661636353, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0153, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1261531263589859, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.0155, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.12295454740524292, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0162, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.14834828674793243, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0162, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.14458344876766205, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0146, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.10006255656480789, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0183, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11527495086193085, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0174, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.11176037788391113, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0146, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.11118462681770325, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0139, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.1216764822602272, + "learning_rate": 1.713029611235338e-05, + "loss": 0.018, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10572952032089233, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0155, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.09299932420253754, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0193, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.11654805392026901, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0146, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.125601127743721, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0148, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.12143341451883316, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0144, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.10944563895463943, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0188, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.10531438142061234, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.017, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.08527344465255737, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0134, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.12489438056945801, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0184, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.12359192967414856, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0162, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.09504237025976181, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0147, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.1144440770149231, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0155, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.11942129582166672, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.0161, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.15496809780597687, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0167, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.11641886830329895, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0155, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.13092041015625, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0164, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.1662251204252243, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0169, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.1151951253414154, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0146, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.13008327782154083, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.017, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.08816810697317123, + "learning_rate": 1.70484286724112e-05, + "loss": 0.0147, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.12602686882019043, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0167, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10360333323478699, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0161, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.10288142412900925, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0174, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.1306314915418625, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0178, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.11859587579965591, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0152, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.12666994333267212, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.0155, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.10859639197587967, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0174, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.1423543244600296, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0145, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.12444380670785904, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0165, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.07491649687290192, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0174, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.12173314392566681, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0156, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.11790800094604492, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.1375807374715805, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0182, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.11894135922193527, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0149, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.09377862513065338, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0157, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1448003053665161, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0189, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.14296171069145203, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.019, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.10862060636281967, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0188, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.1081303209066391, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.0166, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08262617141008377, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0163, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12013579159975052, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0161, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.06014449894428253, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0166, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09688957780599594, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0164, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.11730696260929108, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.0142, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.11243025213479996, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0134, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11974845081567764, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0156, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.1408282220363617, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0193, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.1495579332113266, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.017, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.1408172994852066, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.19624975323677063, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0165, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09940195083618164, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0168, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.11277889460325241, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0171, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.13027498126029968, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0165, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.11570801585912704, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.0159, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.12621626257896423, + "learning_rate": 1.690501241902545e-05, + "loss": 0.016, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.13131381571292877, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0146, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.11577112972736359, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0168, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.1240323930978775, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0162, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.17400938272476196, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.0155, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.09539356827735901, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0165, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10108646005392075, + "learning_rate": 1.688040861067922e-05, + "loss": 0.016, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.1295158863067627, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0151, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.13638637959957123, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0188, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.12255847454071045, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0148, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.13242852687835693, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0166, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11082842200994492, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.015, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.14538930356502533, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0172, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1609591394662857, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0166, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.10611626505851746, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.0142, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.13291358947753906, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0179, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.1112443208694458, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0179, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.10251478105783463, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0174, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.10443990677595139, + "learning_rate": 1.683118563837068e-05, + "loss": 0.016, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.08651736378669739, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0176, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.09996092319488525, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.017, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.13269121944904327, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.0154, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.121453657746315, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0167, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.10587169975042343, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0161, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.12990498542785645, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0158, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.12159401923418045, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0158, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.12046501785516739, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0165, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.10626068711280823, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0144, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.14602169394493103, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0161, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09537681937217712, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0188, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.1014745756983757, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0171, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.11358807235956192, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0165, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.12290225178003311, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0153, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.11178604513406754, + "learning_rate": 1.676962882515376e-05, + "loss": 0.0175, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.1464385688304901, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0155, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.12778854370117188, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0144, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.09556689858436584, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0171, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.10866247862577438, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.0151, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.14120467007160187, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0163, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.12688933312892914, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0174, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.10251957923173904, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0171, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.11465741693973541, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.016, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.08181405812501907, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0201, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.09379345923662186, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0172, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.12122894823551178, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0183, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.0956249088048935, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0142, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.11953603476285934, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0169, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.12494538724422455, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0167, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.20808497071266174, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0138, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.10776039958000183, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0152, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.09606330841779709, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0154, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.10988017171621323, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0182, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.10328707098960876, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0146, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.166867196559906, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0179, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.11701484769582748, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0159, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.10510280728340149, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0179, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.12857040762901306, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0172, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1627577394247055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0176, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.10962127149105072, + "learning_rate": 1.666696746039845e-05, + "loss": 0.0147, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.12961529195308685, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0164, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.08055070787668228, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0148, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.10041354596614838, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0161, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.12196766585111618, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0138, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.0879533439874649, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0156, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.1102190688252449, + "learning_rate": 1.664231677578745e-05, + "loss": 0.016, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.12463485449552536, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.016, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.10829959809780121, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0181, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.08602527529001236, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0176, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.10241527855396271, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.0172, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11007986962795258, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0156, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13110534846782684, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.018, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1417531818151474, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0161, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.13233700394630432, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0147, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.09148544073104858, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0145, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.08781567215919495, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.0149, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.10577751696109772, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0161, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.11689155548810959, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0171, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.0777219608426094, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0152, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.07983718812465668, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0158, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1595761626958847, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0168, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.1037725880742073, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0161, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.10239970684051514, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0149, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.15451422333717346, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0199, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.11882807314395905, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0165, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.09850180149078369, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0147, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.10899021476507187, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.09254994243383408, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0148, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.13559140264987946, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0162, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.11541012674570084, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0154, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.07937203347682953, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0153, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.10062690824270248, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0165, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.13341091573238373, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0178, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.0857294499874115, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0132, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.10551585257053375, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0142, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.12363193184137344, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0165, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.11550596356391907, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0165, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.09379533678293228, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0154, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.10079154372215271, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0163, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.13103018701076508, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0178, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.1172979474067688, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0152, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.11038312315940857, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0152, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.0925338938832283, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0148, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.2519476115703583, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.017, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.09703299403190613, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0193, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.10749004036188126, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0175, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.08736859261989594, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0147, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.13125599920749664, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0161, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10677823424339294, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0156, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.09135954827070236, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0151, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.10548236966133118, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0179, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.1150064468383789, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0188, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.127374529838562, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.016, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.09152884781360626, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0177, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.1021062359213829, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0145, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.1054648831486702, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0172, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.12128820270299911, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0165, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.11375363171100616, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0151, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.11004973948001862, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0182, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.11101444810628891, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0164, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.10393026471138, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0172, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.1139075756072998, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0159, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1569976508617401, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0178, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09649650752544403, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0165, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.16198694705963135, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0157, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.10348767787218094, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0173, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.12259576469659805, + "learning_rate": 1.639145830609088e-05, + "loss": 0.0171, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.10480271279811859, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0159, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.11254257708787918, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0163, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.11534930765628815, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0149, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.10350872576236725, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0169, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.10812029242515564, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.015, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.11796360462903976, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.0144, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.1276526153087616, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0157, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.11665219813585281, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0151, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.10940699279308319, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0165, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.09596745669841766, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0149, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15527229011058807, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0164, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11651212722063065, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0145, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.12004360556602478, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0153, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.13298730552196503, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0176, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.1496417224407196, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0153, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.11104541271924973, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0158, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.12215209007263184, + "learning_rate": 1.632147383062092e-05, + "loss": 0.015, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.09335075318813324, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0148, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.11481857299804688, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0153, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14369069039821625, + "learning_rate": 1.630912057457699e-05, + "loss": 0.019, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.11009083688259125, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0162, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.11112431436777115, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.0137, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.10135089606046677, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0167, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09966760873794556, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0168, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12062226980924606, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0145, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.1109282448887825, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0162, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.10862662643194199, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0152, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.09910363703966141, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0156, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.1374131590127945, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0167, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.12114953994750977, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.0156, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.14391179382801056, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0156, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.12364984303712845, + "learning_rate": 1.625969868712804e-05, + "loss": 0.0169, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.08771048486232758, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0142, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.10959397256374359, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0176, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.138074591755867, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0149, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.1116161122918129, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0154, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.1398090422153473, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0164, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.12145344167947769, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0173, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.11162104457616806, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0153, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.1211397722363472, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.016, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09488996863365173, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0146, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.11799386143684387, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0137, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.1406150758266449, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.0149, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.1307811290025711, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0192, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.13287024199962616, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.0153, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10396360605955124, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0133, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.11891327053308487, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0167, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.12922705709934235, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0178, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10563277453184128, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0132, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.17530488967895508, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0159, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.10669656842947006, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0166, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.09745346754789352, + "learning_rate": 1.617729852865125e-05, + "loss": 0.0152, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.12519428133964539, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0136, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.10747594386339188, + "learning_rate": 1.616905650452066e-05, + "loss": 0.0147, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.10352121293544769, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0161, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.1046888679265976, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0131, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.2108439952135086, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0148, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.1401670128107071, + "learning_rate": 1.61525713933778e-05, + "loss": 0.0171, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.10765372961759567, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0169, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.14557960629463196, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0151, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.11173100024461746, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0181, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.08689062297344208, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0156, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.10052058845758438, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0171, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.11383800208568573, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0167, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10390642285346985, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0157, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.10242070257663727, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0147, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.10699810087680817, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.0162, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.1349048614501953, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0167, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.10787097364664078, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.0144, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.06869964301586151, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0168, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.08449438959360123, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.0156, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10576694458723068, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0157, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.10042312741279602, + "learning_rate": 1.609073991709496e-05, + "loss": 0.018, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.10223086178302765, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0179, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.11301153153181076, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0151, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.11257045716047287, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0159, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.09129524976015091, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0154, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.10601308941841125, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0157, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.11378258466720581, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0151, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.1138806939125061, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0188, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.10367786139249802, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0157, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.13393838703632355, + "learning_rate": 1.605363201758077e-05, + "loss": 0.016, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.09532077610492706, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0145, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.13368268311023712, + "learning_rate": 1.60453849294712e-05, + "loss": 0.0178, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10675307363271713, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0148, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.10722523182630539, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0167, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.09825004637241364, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0148, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.14284753799438477, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.0194, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0167, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.13374702632427216, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0153, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.08964364975690842, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.0164, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.07354595512151718, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0138, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.08927512913942337, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0167, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.10888224840164185, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0193, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.12273052334785461, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0157, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.10807470232248306, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0152, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.1124584972858429, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.0148, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.10052672028541565, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0145, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.08666819334030151, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0172, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.11364969611167908, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0155, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.11263672262430191, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0164, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.1379120796918869, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0187, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.10857756435871124, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0133, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.13408084213733673, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0156, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.10847405344247818, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0141, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.10151072591543198, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0138, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11481117457151413, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0138, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.10202853381633759, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0153, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.09370319545269012, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.0159, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.09684473276138306, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0162, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.11064363270998001, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0162, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.15720140933990479, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0167, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.12132807075977325, + "learning_rate": 1.592576773310894e-05, + "loss": 0.0159, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.10888595134019852, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0133, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.0892426148056984, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0163, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12278196215629578, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0186, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10831111669540405, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0155, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.08722463995218277, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0156, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.11901626735925674, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0155, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.08866825699806213, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.0155, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.09303046017885208, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0152, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.12453419715166092, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0166, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.11046329140663147, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0155, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.10482461750507355, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0154, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11235640943050385, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0165, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.11749114841222763, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0159, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.1201302632689476, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0174, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.13044941425323486, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0143, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.10600301623344421, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0155, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.12556208670139313, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0159, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.12454571574926376, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0153, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.13055634498596191, + "learning_rate": 1.584736531020125e-05, + "loss": 0.0188, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.1379219889640808, + "learning_rate": 1.584323820456826e-05, + "loss": 0.016, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.13937926292419434, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0181, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.21694405376911163, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.013, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.11387006938457489, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0144, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.10102641582489014, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0154, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.10343687981367111, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0161, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.14018288254737854, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0175, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10792441666126251, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0168, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.10692388564348221, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0175, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.11248136311769485, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0176, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.13729289174079895, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0154, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.1389625370502472, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0152, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11402085423469543, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0154, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.14007720351219177, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0151, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.10529543459415436, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0151, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.09746415913105011, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0136, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1150866374373436, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0158, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.10207786411046982, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0139, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.13951019942760468, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0157, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11709082126617432, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.0158, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.12401671707630157, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0145, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.103446826338768, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0149, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.16494140028953552, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0192, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.10893384367227554, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0141, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.09813039004802704, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.0142, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.14741374552249908, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0184, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.12035968154668808, + "learning_rate": 1.573591189751e-05, + "loss": 0.0157, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.0970459058880806, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0176, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.10946980863809586, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0144, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.15628819167613983, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0157, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.12485401332378387, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0184, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.11361467838287354, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0163, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.19872187077999115, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0152, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.0912356749176979, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0167, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.09560416638851166, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0178, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.10187338292598724, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0127, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.11934805661439896, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0145, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.11719757318496704, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0148, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09480439126491547, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0167, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.09031171351671219, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0163, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.12236128002405167, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.014, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.09773887693881989, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.015, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.11999719589948654, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.0145, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.11341046541929245, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0142, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.10415472090244293, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0183, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.10177286714315414, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0152, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.10778342932462692, + "learning_rate": 1.565332720308093e-05, + "loss": 0.016, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.08542678505182266, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0157, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.10727313905954361, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0164, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.1413055658340454, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0163, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.10777363926172256, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0175, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.14286576211452484, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0155, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.10413800925016403, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0164, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.10655509680509567, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0163, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.10989648103713989, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0153, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.09240605682134628, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0144, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.0842762365937233, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0143, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.13544097542762756, + "learning_rate": 1.560789700123175e-05, + "loss": 0.017, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11756332218647003, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0159, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.11726696044206619, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0165, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.11693235486745834, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0146, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.1371903419494629, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0145, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.0971926674246788, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.017, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.08220933377742767, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0156, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.10194069892168045, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0134, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.08653450012207031, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.0131, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.08990923315286636, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0144, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.10370788723230362, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0133, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.06916087865829468, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.0139, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.1136801689863205, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0146, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12992094457149506, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.0144, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.09605134278535843, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0165, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.10273651778697968, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0165, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.11571857333183289, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0151, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.10102164000272751, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0146, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.11045269668102264, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.0146, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.12626338005065918, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0166, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.1357768476009369, + "learning_rate": 1.552528261458322e-05, + "loss": 0.018, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.10587792843580246, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0143, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.1509760171175003, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0187, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11034028977155685, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0174, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.07888679206371307, + "learning_rate": 1.550875774213821e-05, + "loss": 0.0159, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.11219298094511032, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.13006654381752014, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.0161, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12485881894826889, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0148, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.10320480912923813, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0171, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.12367358803749084, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.019, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1412898451089859, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0196, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.10843265801668167, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.0146, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09489759802818298, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0147, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.10236489772796631, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0158, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.12652312219142914, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0173, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10095653682947159, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0167, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.0802559182047844, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0145, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.1239897757768631, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0155, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.09682597219944, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.0154, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.10482888668775558, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.013, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.1272023320198059, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0164, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.09988776594400406, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0166, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.10352584719657898, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0142, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.36612969636917114, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0169, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.12097200751304626, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09437882900238037, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.0147, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14130210876464844, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0147, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.1284940391778946, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0148, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12240798771381378, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0135, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.08258320391178131, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0154, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.10075270384550095, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0162, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.10424498468637466, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0138, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.10447008907794952, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0156, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.10173123329877853, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0127, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.12066177278757095, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0162, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.09489274024963379, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0159, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.09174267202615738, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0155, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.13935479521751404, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0158, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.11126208305358887, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0142, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.13917426764965057, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.014, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.10231155157089233, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0141, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.11477960646152496, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0145, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.10886958241462708, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0155, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.09005019068717957, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0147, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.09915502369403839, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0138, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.1025547906756401, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0158, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.1030225157737732, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0158, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.0842776820063591, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0163, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.0634579285979271, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0151, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.09493366628885269, + "learning_rate": 1.532281505788039e-05, + "loss": 0.014, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12343619018793106, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0167, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.10681836307048798, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0149, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.1302853226661682, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0171, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.10506479442119598, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0136, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.12051218003034592, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0168, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11805758625268936, + "learning_rate": 1.52980183248759e-05, + "loss": 0.0158, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.10309842228889465, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.015, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.108099564909935, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0159, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11023534834384918, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0136, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.11571311205625534, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0163, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.1068359762430191, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.017, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.15622149407863617, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0156, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.11074791103601456, + "learning_rate": 1.526908777801797e-05, + "loss": 0.0163, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.11449336260557175, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0156, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.1219671219587326, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.0167, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.08449103683233261, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0162, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.13202930986881256, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0167, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.11665860563516617, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0173, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.13584332168102264, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0148, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.09913904964923859, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0155, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09704537689685822, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0162, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.13063789904117584, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0153, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.1202530562877655, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0182, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.08398830890655518, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.015, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.09059721976518631, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0129, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.10968997329473495, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0141, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.09898033738136292, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.1568402498960495, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0161, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.09885324537754059, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0168, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09452448785305023, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0142, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.16158591210842133, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0149, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.13292880356311798, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.016, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.17052333056926727, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0165, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.11351195722818375, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0163, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.09296253323554993, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0128, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.09396881610155106, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0157, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.13142332434654236, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0161, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.1246478259563446, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0135, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.13347554206848145, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0173, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.10838479548692703, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.11035843938589096, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0158, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.10406306385993958, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0149, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.16070416569709778, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0149, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09154598414897919, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.014, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.10438812524080276, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0164, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.09018818289041519, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0149, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1320919543504715, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0151, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.0870271846652031, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0155, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.12103381007909775, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0191, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.09007984399795532, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.0141, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.10315215587615967, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.19641031324863434, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0169, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14303788542747498, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0207, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.10933230072259903, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0136, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.11214675009250641, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0175, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.09008116275072098, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0145, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.09065063297748566, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.016, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.1116153672337532, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.016, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.09231071174144745, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0148, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12521710991859436, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0159, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.12240038067102432, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0143, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.09896460920572281, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0154, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.10471562296152115, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0152, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.11139113456010818, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0132, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.09953293949365616, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.015, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.12128512561321259, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0155, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.14064094424247742, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0132, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.09164810925722122, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0144, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.11127574741840363, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0145, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.11218927800655365, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0168, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.10342293232679367, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0137, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.10499373823404312, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0151, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.08029376715421677, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0157, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09445271641016006, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0161, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.09650483727455139, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0155, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.08300688117742538, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0176, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.11890654265880585, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0145, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10451909154653549, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0195, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.08795508742332458, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0141, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.1659373790025711, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0197, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.1001206561923027, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0161, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10190606862306595, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0161, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.11775147914886475, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0139, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.11382468789815903, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0149, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10219351947307587, + "learning_rate": 1.497147766283292e-05, + "loss": 0.0143, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.0930316224694252, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0148, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.10227128118276596, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0153, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.10634683817625046, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0163, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.1443759799003601, + "learning_rate": 1.495494301522528e-05, + "loss": 0.016, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.10867374390363693, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0147, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.11999388039112091, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0157, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10372292995452881, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0161, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.14905284345149994, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.0172, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10412498563528061, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0145, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.09797366708517075, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0133, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.1329793930053711, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0179, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.09716201573610306, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0146, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.09868425875902176, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.0153, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.0929884985089302, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0159, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.1417207568883896, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0151, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.13194335997104645, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0177, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.11850415170192719, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0173, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08122166991233826, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0149, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.11195199191570282, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.12405557930469513, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0135, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.12344473600387573, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0153, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.11776174604892731, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.0154, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.11634577810764313, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0158, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.11508087068796158, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0165, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.12009026855230331, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0141, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.15844415128231049, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.014, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.11803163588047028, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.014, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.11727108061313629, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.0157, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.09923852235078812, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0153, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.14936032891273499, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0174, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.10241764038801193, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0164, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.11330186575651169, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.016, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.14860020577907562, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0152, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.1636815220117569, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.015, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.11093862354755402, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0142, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.11333172768354416, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0141, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.10221018642187119, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0155, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.11085604131221771, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0158, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.17324623465538025, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0153, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.10631918907165527, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.0136, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.09826772660017014, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0148, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.1288551241159439, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11162615567445755, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0148, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.1307016909122467, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.017, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11215298622846603, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0154, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.09471828490495682, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0145, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10456835478544235, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.0147, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.08179501444101334, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0129, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.11810965836048126, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0151, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.0977158397436142, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0139, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.09650871902704239, + "learning_rate": 1.476067039948277e-05, + "loss": 0.0138, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.10262744128704071, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0145, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11392423510551453, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0147, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11900563538074493, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0163, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.10666237771511078, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0162, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.10767115652561188, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0152, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.1232513040304184, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0144, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.12271980196237564, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0167, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.12620051205158234, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0176, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11799243837594986, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0153, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.11633779853582382, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0154, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.10304611176252365, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0173, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.11186012625694275, + "learning_rate": 1.471107401829042e-05, + "loss": 0.0143, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.12054259330034256, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0174, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.1436678171157837, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0146, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.12709002196788788, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0171, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.11617328971624374, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0146, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.1117766946554184, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0178, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.12944459915161133, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0155, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11261707544326782, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0144, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.11154191195964813, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.013, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.11878378689289093, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0142, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.12532871961593628, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0176, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.10492552816867828, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0165, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.10776105523109436, + "learning_rate": 1.466148079674631e-05, + "loss": 0.016, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.10712877660989761, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0165, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.091713547706604, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0155, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.10679589211940765, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0185, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.11124882102012634, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0141, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.10930055379867554, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0148, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.10555030405521393, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0152, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.0836755633354187, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0153, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.11849679052829742, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0158, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.13186359405517578, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.0167, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.11089541018009186, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0189, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.09473541378974915, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0155, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.1151607483625412, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0138, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.10416971892118454, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0138, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.10269354283809662, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0158, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11249537020921707, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0152, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.14352446794509888, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0169, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10745783895254135, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0159, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.09522922337055206, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0164, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.13701501488685608, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0146, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.08020906895399094, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0139, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09878123551607132, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.015, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.09399155527353287, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.015, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.11354538053274155, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0175, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.09135859459638596, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0156, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.08270548284053802, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.014, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.11252828687429428, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0157, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.0985439196228981, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0163, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.12125496566295624, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0146, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.09888023883104324, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0176, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.10250411182641983, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.10701503604650497, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0151, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11525820940732956, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0157, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.09755205363035202, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0159, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09472718834877014, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.015, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10360235720872879, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.0149, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.09946251660585403, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0164, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.12811636924743652, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0135, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.127360999584198, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.016, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.12480203062295914, + "learning_rate": 1.45003312024933e-05, + "loss": 0.0142, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.10674259066581726, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0156, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.12259232252836227, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0149, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.08998828381299973, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0149, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.10316051542758942, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0145, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.14545974135398865, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0158, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.1868322491645813, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.015, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.12345045059919357, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0154, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.11794525384902954, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0177, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.12257082760334015, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0152, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.12155821174383163, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0159, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12235453724861145, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0147, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.1056676059961319, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.015, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.1309097409248352, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0145, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.11743862181901932, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0152, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.12476198375225067, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.016, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.08397012203931808, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0162, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12207506597042084, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0159, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.09591932594776154, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0137, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.12573577463626862, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0164, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1315455585718155, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0164, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.1468929648399353, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0165, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.15008211135864258, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0148, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.09235277771949768, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0153, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.1275569051504135, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0163, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.08151543140411377, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0129, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10390543937683105, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.015, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.11043227463960648, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0141, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.1271016001701355, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0163, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09329107403755188, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0134, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.10737643390893936, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0135, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12903659045696259, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0159, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.11276338994503021, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0147, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.14026683568954468, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0163, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.12503503262996674, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0172, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.10071003437042236, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0138, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10627497732639313, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0138, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.0770413801074028, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0135, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.09400150179862976, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0154, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.1549134999513626, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.015, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.09969423711299896, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.0135, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.09770094603300095, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0149, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.0848713293671608, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0128, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08489260822534561, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0142, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.14606203138828278, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.0144, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.1283080279827118, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0147, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.09238846600055695, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0154, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.09167003631591797, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0153, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.1081148013472557, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0142, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.19600129127502441, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0158, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.08589469641447067, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0148, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10786846280097961, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0123, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.12004593759775162, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0154, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10496815294027328, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0133, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.1324186772108078, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0152, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.10160262882709503, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0137, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.14738032221794128, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0167, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.1030116155743599, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0146, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.10793277621269226, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0156, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.11900396645069122, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0165, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.12115667760372162, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0144, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.12846702337265015, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0163, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.09553872793912888, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.015, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.12807151675224304, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0161, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.12759897112846375, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.017, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.10901748389005661, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0151, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.09706093370914459, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0129, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10346078872680664, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.0147, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.12170220911502838, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0136, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.24494236707687378, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0139, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.10071292519569397, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0131, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.13647668063640594, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0151, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09296777099370956, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.0147, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.1015978679060936, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.0147, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.09014145284891129, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0134, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.08798721432685852, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0136, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.08899430930614471, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.014, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.10661277174949646, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.0189, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.10300924628973007, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.0165, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.10107792913913727, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0152, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.12540587782859802, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0147, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.14788442850112915, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0154, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.1628517359495163, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0155, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.12338890135288239, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0156, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09656467288732529, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0132, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.15762092173099518, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0142, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.1834212690591812, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0156, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.091162770986557, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.0142, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.14656929671764374, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0163, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.08447254449129105, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0144, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10864629596471786, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0137, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.12742409110069275, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0134, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.09852556139230728, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.13655197620391846, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0143, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.09299396723508835, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0134, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.08303989470005035, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0136, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.09860356897115707, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.016, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.08682689070701599, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0134, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.09411480277776718, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.0149, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.10258188843727112, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.016, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.09782259166240692, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0165, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.10300199687480927, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0148, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.09208223968744278, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0175, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.09144087135791779, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0148, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.09090441465377808, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0153, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.10571702569723129, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0168, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.1070566177368164, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0154, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.12103649973869324, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.015, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.10678276419639587, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0165, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11873649060726166, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0154, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.11791808903217316, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0145, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.0986858382821083, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.016, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09249810129404068, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0127, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12782645225524902, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0154, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.10502453148365021, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0162, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.09005366265773773, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.0143, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11655479669570923, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0158, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.10148292779922485, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0155, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.11018893122673035, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0147, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.11230522394180298, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0148, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.09911413490772247, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0159, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.09358150511980057, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0154, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.12689724564552307, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0145, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.10737589001655579, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.013, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.12670603394508362, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0148, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.1459159404039383, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0162, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.14686602354049683, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0139, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10709210485219955, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0161, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.08311928808689117, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.015, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12497760355472565, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0161, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.08203094452619553, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0129, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.12578807771205902, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0145, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10697296261787415, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0147, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.10980327427387238, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0154, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.09966560453176498, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0126, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.09740720689296722, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0126, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.1452278196811676, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.013, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.10068526864051819, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0152, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.10273153334856033, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0134, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.1101747378706932, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0148, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.11141166090965271, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0133, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1535777896642685, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0155, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.08465548604726791, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0161, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.07614064961671829, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0149, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09852079302072525, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0135, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.13796597719192505, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0163, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1669035404920578, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0143, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09881193935871124, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0142, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.12325254082679749, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0132, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.16859179735183716, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0168, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.09475059062242508, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.014, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.11981360614299774, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0186, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.10775262862443924, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0151, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.10053553432226181, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0129, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.10618098825216293, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0154, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.12225139886140823, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0171, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.07435912638902664, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0152, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.10211736708879471, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0177, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12564173340797424, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.015, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.10412495583295822, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0157, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.12016817182302475, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0146, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.10957430303096771, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0141, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08035340160131454, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0146, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.1221131756901741, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0153, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.10435835272073746, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0133, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.11107036471366882, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0137, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.0575222373008728, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0149, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.076436348259449, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0159, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.09325376152992249, + "learning_rate": 1.380704261284659e-05, + "loss": 0.0163, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.11715361475944519, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0161, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.08567722141742706, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0149, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.12162303924560547, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.017, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.11919409781694412, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.0154, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.10694293677806854, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0138, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.10736183077096939, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0133, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11603212356567383, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0127, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.1180153489112854, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.0161, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.11968789249658585, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.0151, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.1250530481338501, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0136, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10276933014392853, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0174, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.14728401601314545, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.0138, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.09268351644277573, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0129, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12262973934412003, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0147, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.11068148910999298, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0157, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.09095525741577148, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0149, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.12430994212627411, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0154, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1246076300740242, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0153, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.08052714169025421, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.0144, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.1667853593826294, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0164, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.10500273108482361, + "learning_rate": 1.372053087451907e-05, + "loss": 0.013, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.10832768678665161, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0171, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.11245214939117432, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0158, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.14754413068294525, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0146, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.16368411481380463, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0175, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.1533215045928955, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0138, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.11878912895917892, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0134, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.10748414695262909, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0155, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.14132313430309296, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0139, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.09151305258274078, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0133, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10014014691114426, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0143, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.12393905967473984, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0154, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13815879821777344, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0155, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.11488667875528336, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.014, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.08504864573478699, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0149, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.12521564960479736, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0161, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.09525968134403229, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0137, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.12103450298309326, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.017, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.10642504692077637, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0139, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.15175847709178925, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0153, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11594653874635696, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.017, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.12788720428943634, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0171, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.12372581660747528, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0157, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.09206387400627136, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0155, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.09841199964284897, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0162, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.135525643825531, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0148, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.0982549637556076, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.0148, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.07432328909635544, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0124, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09830821305513382, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0139, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.10874123871326447, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0149, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.08979356288909912, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0189, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.08131087571382523, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.0128, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.1204582154750824, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0141, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11884505301713943, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0144, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.16451388597488403, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0134, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.11953870207071304, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0135, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12111926078796387, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0159, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.1181047335267067, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0142, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.12142453342676163, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0156, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09890097379684448, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.0137, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.13294243812561035, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0132, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.1397709846496582, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0161, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.10867296159267426, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.014, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11157155781984329, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.015, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09516428411006927, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0156, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.21431738138198853, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0158, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.09854292124509811, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0159, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.21810708940029144, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0152, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.14076393842697144, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0162, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.0892275720834732, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0154, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.09736334532499313, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0127, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.13880307972431183, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.015, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1091906875371933, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0147, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11605603992938995, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0168, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.11140985041856766, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0162, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.09057680517435074, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0161, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.12626419961452484, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0161, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.136195570230484, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0142, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.12415066361427307, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0121, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.12581513822078705, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.0151, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.11835133284330368, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0147, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.12255530059337616, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.0135, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10576335340738297, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0149, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.12575241923332214, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.014, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09812266379594803, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0152, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.11845924705266953, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0143, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.10975389182567596, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0147, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.09649760276079178, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0131, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.12912778556346893, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.0152, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10225717723369598, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0143, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.1712195724248886, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0144, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.1477300226688385, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0155, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.11692417412996292, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0146, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.1232081949710846, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0152, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.14507624506950378, + "learning_rate": 1.341604094324139e-05, + "loss": 0.017, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.13909533619880676, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0136, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.1438206136226654, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.0167, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.09637414664030075, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0134, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.15815401077270508, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.015, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.12961018085479736, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0159, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.11171260476112366, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0133, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.0907273143529892, + "learning_rate": 1.33872699666556e-05, + "loss": 0.015, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.11181363463401794, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0162, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.11502614617347717, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0169, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.14428892731666565, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0158, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.1440858244895935, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0134, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.07517627626657486, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0129, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.11884543299674988, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0147, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.08753574639558792, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0178, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.09831796586513519, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.014, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.12262003868818283, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.016, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.11444123834371567, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0142, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11559469252824783, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0138, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.09936029464006424, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0116, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.12810789048671722, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.0137, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.07923603802919388, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0136, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.09709552675485611, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0148, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.0924353376030922, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0152, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.10444460809230804, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0136, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.09510108828544617, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.014, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.11014958471059799, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.014, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.13198788464069366, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0153, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.09857133030891418, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0128, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.15041643381118774, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0148, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.13857626914978027, + "learning_rate": 1.329277958632556e-05, + "loss": 0.018, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.13004948198795319, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0142, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.09864714741706848, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0143, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12682363390922546, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0177, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.09303848445415497, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0141, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.13559986650943756, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0153, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.10143767297267914, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0131, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.1287422627210617, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0142, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.12000516802072525, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.016, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.10434254258871078, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.014, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08141680061817169, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0138, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08781418204307556, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0138, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.1355503350496292, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.015, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.1243978887796402, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0143, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.11088231950998306, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0154, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12559033930301666, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0132, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.16149592399597168, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.09809687733650208, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.0142, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.12586510181427002, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0165, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.12403467297554016, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.14458157122135162, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.0135, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.09313284605741501, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0165, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.14337806403636932, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0144, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.12650802731513977, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0126, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.112495556473732, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.0154, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11324562877416611, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.0139, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.11720716208219528, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.0146, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.1261121928691864, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0143, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.1302812099456787, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0157, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.12074442952871323, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0172, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.11716746538877487, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0158, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.11565797030925751, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0144, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.11653099209070206, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0155, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.13103044033050537, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0152, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.10341408103704453, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0138, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.11098334193229675, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.0146, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.11139992624521255, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.10786942392587662, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0143, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.10139934718608856, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0154, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.12619639933109283, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0153, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.13168810307979584, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0154, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.09545969218015671, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0151, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.09984135627746582, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0129, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10903751850128174, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0135, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.11882342398166656, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0151, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.1105988621711731, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0137, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.09874245524406433, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0142, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.10673530399799347, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0157, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.10264459252357483, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0162, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10769633203744888, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0133, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.16478870809078217, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0149, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.12630558013916016, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0128, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09824837744235992, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0144, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12136032432317734, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0124, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.10943865031003952, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0154, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.11394404619932175, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.0117, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.10907739400863647, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0144, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10671219974756241, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0156, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.10080806165933609, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0141, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.17395640909671783, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0145, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.12620997428894043, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0186, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.13434381783008575, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0149, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.09921885281801224, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.0149, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.13742485642433167, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0155, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.0998004674911499, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10389069467782974, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0138, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.11996421217918396, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.0147, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.09136862307786942, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0129, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.11111150681972504, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0159, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.11024610698223114, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0138, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.15376223623752594, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0154, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.13934020698070526, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0138, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.11208872497081757, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0131, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1513150930404663, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.10351566225290298, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0137, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11078765988349915, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0164, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.15622863173484802, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.0147, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.11603175848722458, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0145, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.1254235804080963, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0162, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.0900387316942215, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.015, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09852428734302521, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0172, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.13502047955989838, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.014, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.1460445076227188, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0156, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.10317493230104446, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0129, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.127616748213768, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0157, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.11560963094234467, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0142, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.11239839345216751, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0135, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.1503314971923828, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0157, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.1253356635570526, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0154, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.13217659294605255, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0127, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09087327867746353, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0144, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.09290582686662674, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0147, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.10273045301437378, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0165, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.14741986989974976, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.016, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.0804770216345787, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0125, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.12413053959608078, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0136, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.09822124987840652, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0176, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.09091541171073914, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0161, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.09487786144018173, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0148, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1009742021560669, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0147, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.08731451630592346, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0118, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.08123810589313507, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0134, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.10559827089309692, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0159, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.11313227564096451, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0164, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.11703694611787796, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0171, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.1083052009344101, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0168, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.13762544095516205, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0185, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.07979343086481094, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.09789150208234787, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0139, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.1905684620141983, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0158, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.1036764457821846, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0118, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.09413912892341614, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0137, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.2435447722673416, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.0132, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.09290791302919388, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0146, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.1211535856127739, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0151, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.1133081242442131, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.013, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.11362706869840622, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0171, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.10019481182098389, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0151, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.09489522129297256, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0143, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12912064790725708, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0138, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.11348122358322144, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0136, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.1325748860836029, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0156, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.11978204548358917, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0146, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.10022053122520447, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0137, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.07388417422771454, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0152, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.10844552516937256, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.0142, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.12381294369697571, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0132, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09625408798456192, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.0154, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.1207597553730011, + "learning_rate": 1.276827841111951e-05, + "loss": 0.016, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.12216818332672119, + "learning_rate": 1.276419082873553e-05, + "loss": 0.016, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.0969374030828476, + "learning_rate": 1.27601034161465e-05, + "loss": 0.0125, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.13043035566806793, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0145, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.10600738227367401, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0144, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.11113949120044708, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0135, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.09458228200674057, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0139, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10076881945133209, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0139, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1297597885131836, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0142, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.10059160739183426, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0142, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10716533660888672, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0149, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.1138077974319458, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0151, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.13693730533123016, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0136, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.08157949149608612, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0143, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.09702442586421967, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0121, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09029403328895569, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0133, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.1444508135318756, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0133, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.11026642471551895, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0161, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.11229082942008972, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0135, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.08556455373764038, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0166, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10684676468372345, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.0142, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.11970192193984985, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0132, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.10227229446172714, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.015, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.08301195502281189, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0153, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.1275268793106079, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.014, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.11613821983337402, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.015, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.10491939634084702, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0127, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.09623268246650696, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0136, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.111306332051754, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0137, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0863005742430687, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.0137, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.09434706717729568, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0142, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.10771634429693222, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0155, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.14157304167747498, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.0145, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11419932544231415, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0154, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.12308389693498611, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0133, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.10775505006313324, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.0144, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.10383253544569016, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0141, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.1012682244181633, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0144, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.12071908265352249, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0157, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.12535089254379272, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0149, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.08634085953235626, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0152, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.10035993158817291, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0138, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.13245077431201935, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0135, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.08713636547327042, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0144, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.10049428045749664, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0134, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.10432936251163483, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0163, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.16578327119350433, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0166, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.16106748580932617, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0125, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.11968962103128433, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0142, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1189485490322113, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0155, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.10795646905899048, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0139, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.08580315858125687, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0134, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.17969708144664764, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0159, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09675301611423492, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0146, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.10921840369701385, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0158, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.11278913170099258, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0133, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12176436185836792, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0138, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.1186685562133789, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0171, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.10296562314033508, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.015, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.15394653379917145, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0173, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.06815146654844284, + "learning_rate": 1.252333462101563e-05, + "loss": 0.014, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.1100328117609024, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0129, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.08578243851661682, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0145, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.11262256652116776, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.015, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.09644307941198349, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0183, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09589987248182297, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0145, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.09658264368772507, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0144, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.11963947862386703, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0118, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.14156748354434967, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0164, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.14387111365795135, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0151, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.11542942374944687, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0123, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.12162452191114426, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0165, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.13585466146469116, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0157, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.11272095143795013, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0159, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.1828201562166214, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.0162, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.09920861572027206, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0163, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.12618015706539154, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.016, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.13111864030361176, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0155, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.08486807346343994, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0126, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.14975349605083466, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0156, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.12495499104261398, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0154, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.11145244538784027, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0147, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.08242745697498322, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0149, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.10702498257160187, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0152, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.16176027059555054, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0145, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.0959012433886528, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0134, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10282868891954422, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0143, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.11747556924819946, + "learning_rate": 1.241332681912234e-05, + "loss": 0.0165, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10770756006240845, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0134, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.09925016760826111, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0126, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.09207499027252197, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.016, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.21546097099781036, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0131, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.10848663747310638, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0182, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.09382427483797073, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0134, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.10758832842111588, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0136, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.10075095295906067, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0154, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.09366318583488464, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0151, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.13190370798110962, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0154, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.1926490068435669, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0162, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.10018754750490189, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0127, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10250869393348694, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0146, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.08466067910194397, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0152, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.10397373139858246, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.0148, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14372816681861877, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0151, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11059669405221939, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0133, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09743087738752365, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0151, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.11265920102596283, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0145, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.08748266100883484, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0146, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.10545652359724045, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0133, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.09680107235908508, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0149, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.13201938569545746, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0152, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.12063595652580261, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0167, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.09986719489097595, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0153, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.09913598746061325, + "learning_rate": 1.230752865632506e-05, + "loss": 0.0151, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.1373804211616516, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0144, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.12088140845298767, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0137, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.08200784027576447, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0145, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.12470699846744537, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0154, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.10114862769842148, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.13169868290424347, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0159, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.08321736752986908, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0147, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.10108216106891632, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.012, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.08475375175476074, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0155, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.08946104347705841, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0145, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.07863342016935349, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0143, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.0906776711344719, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0123, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.11063490062952042, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0128, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.10001818835735321, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0133, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.12135739624500275, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0145, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.1233866810798645, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0133, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.09280609339475632, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0161, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.11677120625972748, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0144, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.11756281554698944, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.0163, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.08084359019994736, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.1415908932685852, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0149, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09107301384210587, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.013, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.11712655425071716, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.014, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.08366917073726654, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0125, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10391885787248611, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0146, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.1042654886841774, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0132, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.09504523873329163, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0148, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.12793844938278198, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0164, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.16210904717445374, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.15506179630756378, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.0153, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.11274553835391998, + "learning_rate": 1.218156580082751e-05, + "loss": 0.015, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.14852653443813324, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0172, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13558489084243774, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0156, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.08902081102132797, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0152, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12315832078456879, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0154, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.13528762757778168, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0145, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.1538013368844986, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0151, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.09894628077745438, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0154, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07696802169084549, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0136, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.09229687601327896, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0131, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.11709824949502945, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0146, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09195996820926666, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0142, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.11714771389961243, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.0157, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.11561911553144455, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0162, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.12904389202594757, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0129, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.10347782075405121, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0126, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.10054155439138412, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0126, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.1713176816701889, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0142, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10279975831508636, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0133, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.11513496935367584, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0137, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.096926249563694, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0123, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.1159687489271164, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.0145, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09745503962039948, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.014, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.15608133375644684, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.0144, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.08617804199457169, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.09285406023263931, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0123, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.11964505165815353, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0143, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1376206874847412, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0136, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.12900781631469727, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0148, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.10527875274419785, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0157, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.11285959929227829, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0135, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.09756818413734436, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0158, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.10695865005254745, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09360404312610626, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0129, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.100432388484478, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0126, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.10982918739318848, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0158, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.07767278701066971, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.015, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.10380839556455612, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0153, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.1225915253162384, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0167, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.0981304869055748, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0139, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.10055239498615265, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0146, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.09311894327402115, + "learning_rate": 1.201528720560666e-05, + "loss": 0.016, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.11859868466854095, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0163, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.15336444973945618, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0153, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.09175246208906174, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0125, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.0973501205444336, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0155, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.1643284410238266, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0142, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.1077687218785286, + "learning_rate": 1.19909852066355e-05, + "loss": 0.015, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.07722865790128708, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0125, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.149483323097229, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.015, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.11405838280916214, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0144, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.11132190376520157, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0142, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.10609457641839981, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0128, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.09458910673856735, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0134, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.10288971662521362, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0164, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.0970534011721611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0128, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.09571509063243866, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0129, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.08922260254621506, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0144, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0886709988117218, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.017, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.11732205003499985, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0141, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.11602790653705597, + "learning_rate": 1.193835918571248e-05, + "loss": 0.016, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.11643379926681519, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.0156, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.0927688404917717, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0149, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11533524841070175, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0133, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11001960933208466, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0135, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.12629783153533936, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.014, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.13336558640003204, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0121, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.11672402173280716, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.016, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.10014890134334564, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0136, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.0940401554107666, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.0151, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.1272585242986679, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0139, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.08492123335599899, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0151, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.1195707842707634, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0137, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.1095622330904007, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0167, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.09656022489070892, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0137, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1358073204755783, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0166, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.10981698334217072, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0141, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.07301078736782074, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0127, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.1040695309638977, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0143, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.16375352442264557, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.0136, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.19000792503356934, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0172, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11925465613603592, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0148, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.188803032040596, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.0151, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.09543585777282715, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0129, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11177625507116318, + "learning_rate": 1.184130718987356e-05, + "loss": 0.013, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.11703382432460785, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0153, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.12326977401971817, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0151, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.09018218517303467, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0147, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1641266793012619, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0146, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.08780542016029358, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0127, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09138940274715424, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0133, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.11768870800733566, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0128, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10298353433609009, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0141, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.10553838312625885, + "learning_rate": 1.180494808593063e-05, + "loss": 0.014, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10878341645002365, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0166, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10571706295013428, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0166, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.09349539875984192, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0126, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.10237022489309311, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0128, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.11938492953777313, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0167, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.10417849570512772, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0132, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.10322093963623047, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0141, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.09445950388908386, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.0146, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.1363389790058136, + "learning_rate": 1.176860863606944e-05, + "loss": 0.014, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.10975046455860138, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.013, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.12059775739908218, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0136, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.10529737174510956, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0158, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.11280884593725204, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0146, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.09665497392416, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0126, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.14059090614318848, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0142, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.10380115360021591, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.0132, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.10641331225633621, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0157, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.10845568776130676, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0135, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.10574200004339218, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0153, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.09128011763095856, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0142, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.10704760998487473, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0129, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.1180003359913826, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0141, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.11784005165100098, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0138, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.08203493803739548, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0144, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.10125461220741272, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.0142, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.12113944441080093, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0124, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.09266798943281174, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0156, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.11875846236944199, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0146, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.11041444540023804, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.013, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.13168254494667053, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0151, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.08208412677049637, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0132, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.08591622114181519, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0152, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.14605054259300232, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.014, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.12816093862056732, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0163, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.10751980543136597, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.013, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.1012972742319107, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0134, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13583813607692719, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0121, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.13573803007602692, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0143, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11719135195016861, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.0163, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.11966057121753693, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0133, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.11172359436750412, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0146, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10357552021741867, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0134, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.09571186453104019, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0143, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.13586629927158356, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0162, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.10461284965276718, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0134, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.182363823056221, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.014, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.12570780515670776, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0161, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.10319641977548599, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0136, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1066908910870552, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0123, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.10409434139728546, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0143, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.08230862766504288, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0128, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.110751211643219, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0148, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.09656667709350586, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0137, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.10626243054866791, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0127, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10294152796268463, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0151, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.09571471065282822, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0136, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.1162625253200531, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0129, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.09755981713533401, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0151, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.08586591482162476, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.016, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.11462593078613281, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.013, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.11438485980033875, + "learning_rate": 1.155904354214324e-05, + "loss": 0.016, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.15895701944828033, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0161, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.1135995015501976, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0128, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.11993709206581116, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0142, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.14105743169784546, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0118, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.08521834015846252, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0136, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.17094957828521729, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0159, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.1379796713590622, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.012, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.10417291522026062, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0138, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.09216596931219101, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0151, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.08398481458425522, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0124, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.1291581243276596, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0152, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.09349840134382248, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0127, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.08282706141471863, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0143, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.10378136485815048, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0146, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1504564881324768, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0144, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.0940636470913887, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0147, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.11142407357692719, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0127, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.16776388883590698, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0158, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.10488522052764893, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0132, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.09612040966749191, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0119, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.09318249672651291, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0116, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.08647014945745468, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0124, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.1067274957895279, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0136, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.08162420988082886, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0136, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.08949539810419083, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0136, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.11327249556779861, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0158, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.10963857173919678, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0141, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.10124490410089493, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.014, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10530558228492737, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0135, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.1251840591430664, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0127, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.14418582618236542, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0156, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.09544084221124649, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0137, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.09794268012046814, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0151, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.10741079598665237, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0127, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.10632490366697311, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0117, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08640208840370178, + "learning_rate": 1.141437135633602e-05, + "loss": 0.0119, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.09476958960294724, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0134, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.1141064390540123, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0147, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.1008499413728714, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0119, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.0853564441204071, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.12970876693725586, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0128, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11144055426120758, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0137, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.09443596750497818, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0127, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.19357536733150482, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0173, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.11861733347177505, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0145, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.13152548670768738, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0145, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11607249826192856, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0128, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.09456845372915268, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0145, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.12110379338264465, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.014, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.07872474938631058, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0128, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.1042177751660347, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0136, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.15168064832687378, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0126, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.09023087471723557, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0125, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.1110723614692688, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0121, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.09889500588178635, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0136, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.13798406720161438, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0151, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.14507459104061127, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0175, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10034530609846115, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0143, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.09978580474853516, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0142, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.11770957708358765, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0145, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.10433546453714371, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0148, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.1006259173154831, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.016, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.09567620605230331, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0133, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08685444295406342, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0116, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.10901349037885666, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.0137, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.15114818513393402, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0147, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.08353722840547562, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0146, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.14387859404087067, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0135, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.12125595659017563, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0148, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.11398079991340637, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0142, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.14072677493095398, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0149, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.11925073713064194, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0153, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.10301142185926437, + "learning_rate": 1.126604838049735e-05, + "loss": 0.013, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.11580533534288406, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.012, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.10776577144861221, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0143, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.1724127233028412, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0146, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.15091492235660553, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0128, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.10073890537023544, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0131, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11136317253112793, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0127, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.11012828350067139, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0133, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.12291619181632996, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0148, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.15791606903076172, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0144, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.12378063052892685, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.0137, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12597671151161194, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0172, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.1783512681722641, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0169, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.07375630736351013, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0139, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.07327539473772049, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0125, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.09731314331293106, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0144, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.0931868702173233, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0124, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.10977564752101898, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0132, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.0881611704826355, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0168, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.1239505335688591, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0133, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.10325739532709122, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0132, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11499917507171631, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0119, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.10131840407848358, + "learning_rate": 1.117803938925701e-05, + "loss": 0.014, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.10435229539871216, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0119, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.10805722326040268, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0136, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0759270042181015, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0167, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.1406354457139969, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0155, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1452491581439972, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.10772743821144104, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0148, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09850231558084488, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0155, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.08501581847667694, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0155, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11223053187131882, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0146, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.07870364189147949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0136, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.11768847703933716, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.0153, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.096342071890831, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0137, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.10153007507324219, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0139, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.0837850570678711, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0117, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.11286687850952148, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0148, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.14671675860881805, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0117, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.09201597422361374, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0161, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.11397483944892883, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0127, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.08533219248056412, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0126, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.18251192569732666, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0134, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.17834120988845825, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.014, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11550542712211609, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0127, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11998283863067627, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0149, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11872667074203491, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0155, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.10344526916742325, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.015, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.11288250237703323, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0149, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.16417740285396576, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0128, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09344452619552612, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0131, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11792241781949997, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0159, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10204951465129852, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0143, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.13064894080162048, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0151, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.08712288737297058, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0148, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.11878253519535065, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0123, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0907600000500679, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0135, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.08273347467184067, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0143, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.08222810924053192, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0112, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.11497339606285095, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0166, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.115402452647686, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0123, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.09611446410417557, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0128, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.12478714436292648, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0133, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.1713663637638092, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0118, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.13887059688568115, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0144, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.08571378141641617, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0154, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.1155383437871933, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0172, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.1123170331120491, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0153, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.13517709076404572, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0142, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.10638624429702759, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0121, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.11200621724128723, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0129, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10821404308080673, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.0128, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.10840310156345367, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0119, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.10352504998445511, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0144, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.10084252059459686, + "learning_rate": 1.097058241668254e-05, + "loss": 0.0128, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.11881101131439209, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.09931256622076035, + "learning_rate": 1.09626195543497e-05, + "loss": 0.0148, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.1260150521993637, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0125, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14202995598316193, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0133, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.15601785480976105, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0158, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.10701733827590942, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0139, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10180657356977463, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0138, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.0962422713637352, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.0124, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.13017651438713074, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0153, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.1295875608921051, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0154, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.10589175671339035, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0144, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.10173171758651733, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0142, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11211177706718445, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.0144, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.09875764697790146, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.014, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.11233554035425186, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0135, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.11327439546585083, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0136, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.10059859603643417, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0143, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.1160704717040062, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0132, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.15698356926441193, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0163, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.08235745877027512, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0134, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.12417344003915787, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0145, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.25982916355133057, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0155, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.12248262017965317, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0139, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.11425210535526276, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.015, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.12354899197816849, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0187, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.1102118268609047, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0137, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11539250612258911, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.015, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.12405017763376236, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0136, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.10405998677015305, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.014, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.0918804481625557, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0162, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.10429271310567856, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0128, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.09917982667684555, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0133, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.15390615165233612, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0133, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.11222631484270096, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0151, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.14449205994606018, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0165, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1332763433456421, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.015, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09963493794202805, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0132, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.12338289618492126, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0122, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.09369654208421707, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0127, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.13490226864814758, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0149, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.08246663957834244, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0158, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.0892438217997551, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0124, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.1275915652513504, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0135, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.150055930018425, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0118, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.08317377418279648, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0141, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.118314728140831, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.0131, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.086687833070755, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0128, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.10310143977403641, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0125, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.06559336930513382, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0135, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.11680027842521667, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0135, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.1381155252456665, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.015, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.08212172240018845, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0139, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.11066312342882156, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0145, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.11565965414047241, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0131, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.08814681321382523, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0122, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.11116618663072586, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0156, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.10491067916154861, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0157, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.15800951421260834, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0141, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.10169819742441177, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0156, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.07719215005636215, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0127, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.08958051353693008, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0145, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.1171223521232605, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0141, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.09662648290395737, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0121, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.09235699474811554, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0129, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.127174973487854, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0157, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.10298880189657211, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0146, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08706901967525482, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0132, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.10898775607347488, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0117, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.12178633362054825, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0129, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.10881572961807251, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0138, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.12326113879680634, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.013, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.13331058621406555, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0129, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.11921956390142441, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0154, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.10736659914255142, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.014, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09894116222858429, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0141, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.09651704877614975, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0146, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.11836998164653778, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.0176, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.1108580157160759, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0129, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.1281266212463379, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0138, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.13581599295139313, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0149, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10911770164966583, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0162, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.14131440222263336, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0126, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.10660075396299362, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0144, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.12547564506530762, + "learning_rate": 1.063722589416218e-05, + "loss": 0.0142, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11465640366077423, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0143, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.1274002492427826, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0145, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.10782597959041595, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0144, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.08273816108703613, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0126, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.12750636041164398, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0113, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.10962793231010437, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0132, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11728726327419281, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0146, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.10677216202020645, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0146, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.0974075049161911, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.0129, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.09190504997968674, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0129, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.10864423215389252, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0122, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.08833710849285126, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0141, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.14306439459323883, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.013, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.11269631236791611, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0115, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.08850698918104172, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0116, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10025947540998459, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0116, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.07444360852241516, + "learning_rate": 1.057003955600053e-05, + "loss": 0.013, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.10606181621551514, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0133, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.11394638568162918, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0143, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.12015067040920258, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0128, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.11465302109718323, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0117, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.09164674580097198, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0116, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.10532126575708389, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0128, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.16767282783985138, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0137, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.16740484535694122, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0151, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08568812906742096, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0135, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.24568453431129456, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0138, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.13317807018756866, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0142, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11310465633869171, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0139, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.09677092730998993, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0123, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.13912886381149292, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0152, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.0902378037571907, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0124, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09430436044931412, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0116, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.13361187279224396, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0161, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09588316082954407, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0135, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.13104380667209625, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0161, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.11476746946573257, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0137, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.087586909532547, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0124, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.12136878073215485, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0156, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.09384764730930328, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0137, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.10422758758068085, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0142, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.11391496658325195, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0125, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.12965570390224457, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0137, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.13042569160461426, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.0159, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.10433127731084824, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0129, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.0989370346069336, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0132, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.08399394899606705, + "learning_rate": 1.045171352558859e-05, + "loss": 0.013, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.11693604290485382, + "learning_rate": 1.044777463361479e-05, + "loss": 0.0148, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.12304933369159698, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0121, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.10788783431053162, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0138, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11043704301118851, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0135, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.11343314498662949, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0158, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.10075241327285767, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0143, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.08752909302711487, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.012, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.11036614328622818, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0126, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.09241441637277603, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.0157, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.1369611620903015, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.0132, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.12782007455825806, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0136, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.140136256814003, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0144, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11508572846651077, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.0136, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.14334864914417267, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0138, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.10154326260089874, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0135, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.11290924996137619, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0134, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.13472969830036163, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0144, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.1215982660651207, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0125, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.10628669708967209, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0125, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.07950089126825333, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0135, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.14542941749095917, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0145, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.1189931258559227, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0115, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.11141669005155563, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0135, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.08776746690273285, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0136, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10942544043064117, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0139, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.15493649244308472, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0123, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.1485723853111267, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.0141, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.0886952131986618, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0124, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.09021062403917313, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0132, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.17778685688972473, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0126, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.11892827600240707, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0125, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.14066243171691895, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0135, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.10870271921157837, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0147, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.10636188089847565, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0156, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.11178954690694809, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.014, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.09059949964284897, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.0127, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.11947090923786163, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0145, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.1133369505405426, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.014, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.12451290339231491, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0139, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10219574719667435, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0138, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.09843198210000992, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0144, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.09784907102584839, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0155, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1259097307920456, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0156, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.08604080229997635, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0113, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.11517608165740967, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0131, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.09721313416957855, + "learning_rate": 1.027088684228894e-05, + "loss": 0.011, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.08404546976089478, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0164, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.11207510530948639, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0151, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.1101691946387291, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0129, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.2047414630651474, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0143, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.08887176960706711, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0139, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.08608511090278625, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0162, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10840177536010742, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0133, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.1228431761264801, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.0173, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.09171874076128006, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0119, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.1357734203338623, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0151, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.13374143838882446, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0134, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09826765209436417, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0112, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.30271241068840027, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0128, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.1474897861480713, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.0142, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.11385788768529892, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0121, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.1145087331533432, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0125, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.15277381241321564, + "learning_rate": 1.020425036775911e-05, + "loss": 0.011, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10591624677181244, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0123, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.09938671439886093, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0136, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.12457235157489777, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0145, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.10550948232412338, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0127, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.12241430580615997, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.013, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.12996822595596313, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0127, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.10886852443218231, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0124, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.10862760245800018, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0124, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.12071339040994644, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0152, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.12002567946910858, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0144, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.0731884092092514, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.014, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.09087152034044266, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0122, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.14147141575813293, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0126, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.10024842619895935, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.013, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.12605364620685577, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0129, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.07398458570241928, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.012, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.09070646017789841, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0139, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.06930165737867355, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.012, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.10134197026491165, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0131, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.12521374225616455, + "learning_rate": 1.012598938117799e-05, + "loss": 0.012, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.10238701850175858, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0139, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.10689418017864227, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0138, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.0954018235206604, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0125, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.1394306868314743, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0125, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.11217380315065384, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0143, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.1513020396232605, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.0122, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.11155983060598373, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0132, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.1257336288690567, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.015, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.10818599164485931, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0163, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.15947701036930084, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0142, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.10929335653781891, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0129, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.08994319289922714, + "learning_rate": 1.007910370568197e-05, + "loss": 0.011, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.09995368868112564, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0131, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.10167150944471359, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.0142, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1480625420808792, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0129, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.11566069722175598, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.013, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.10494371503591537, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0132, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.09531164914369583, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0116, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.0887216180562973, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0133, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.09848874062299728, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0122, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.14705902338027954, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0135, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.1412736475467682, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11742296814918518, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0143, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.11000296473503113, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0144, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.111237533390522, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0125, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.1085008904337883, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0133, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.12815521657466888, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0138, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.11128585785627365, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.12439922988414764, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0139, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.11863227933645248, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0154, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.13517265021800995, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0128, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.0960967019200325, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0123, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.0953422263264656, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0121, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.1965440809726715, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.0924379900097847, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0112, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.08345641940832138, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0146, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.10248678177595139, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0121, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.2980184257030487, + "learning_rate": 9.977703371038337e-06, + "loss": 0.013, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.11192528903484344, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0133, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.21329663693904877, + "learning_rate": 9.969913958481688e-06, + "loss": 0.0116, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.0922258049249649, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0159, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.12964527308940887, + "learning_rate": 9.962126073932371e-06, + "loss": 0.0143, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.11362354457378387, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0137, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.0902123972773552, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0126, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.08770131319761276, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0123, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.09604862332344055, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0131, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.08182749897241592, + "learning_rate": 9.942663063115248e-06, + "loss": 0.013, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.17937488853931427, + "learning_rate": 9.938771611983594e-06, + "loss": 0.0158, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.1343632936477661, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0139, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.14316903054714203, + "learning_rate": 9.93098986311647e-06, + "loss": 0.012, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11366818100214005, + "learning_rate": 9.927099565971976e-06, + "loss": 0.0129, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0949450209736824, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0144, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.13775621354579926, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0134, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.10221140086650848, + "learning_rate": 9.915430987239967e-06, + "loss": 0.012, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.11101876199245453, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0121, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.12785188853740692, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0146, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.1400655061006546, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0115, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.09468253701925278, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0122, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.08724251389503479, + "learning_rate": 9.895991082746168e-06, + "loss": 0.0143, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.10251675546169281, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.11980220675468445, + "learning_rate": 9.888217832192542e-06, + "loss": 0.015, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.12473925948143005, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0144, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.08260339498519897, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0122, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.11015481501817703, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0154, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1120007336139679, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0143, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09239251911640167, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0146, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10413528233766556, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0144, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.0863887220621109, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0143, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.0963623896241188, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0129, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.2090010792016983, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0155, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12460873275995255, + "learning_rate": 9.84937491900087e-06, + "loss": 0.012, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.11049462109804153, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0134, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.15180432796478271, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0142, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.13392038643360138, + "learning_rate": 9.837729659166156e-06, + "loss": 0.013, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.11534144729375839, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0134, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.1077096238732338, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0135, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.10903865098953247, + "learning_rate": 9.826087927694543e-06, + "loss": 0.014, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.09442178905010223, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0132, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.12171642482280731, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0143, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.08623645454645157, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0135, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.10249772667884827, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0151, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.16136185824871063, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0113, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.15169787406921387, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0161, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.095759317278862, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0143, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.11120729893445969, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0127, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09804334491491318, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0123, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.11462689191102982, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0111, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10223866254091263, + "learning_rate": 9.783431894635035e-06, + "loss": 0.012, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.1001039668917656, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0149, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.13274405896663666, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0136, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.09141503274440765, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0135, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.1291927844285965, + "learning_rate": 9.767932474193758e-06, + "loss": 0.0136, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10977526009082794, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0135, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.1167440190911293, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0138, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.08445189893245697, + "learning_rate": 9.756312079913248e-06, + "loss": 0.0129, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.2519867718219757, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0146, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.08974999934434891, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0152, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.1252872496843338, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0147, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.16116158664226532, + "learning_rate": 9.740823797373594e-06, + "loss": 0.013, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.08403292298316956, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0119, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.11238662898540497, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0154, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10106278210878372, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0136, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.08145973831415176, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0131, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.15683840215206146, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0132, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.14468176662921906, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0146, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.08105593919754028, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0122, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.12362894415855408, + "learning_rate": 9.709866422291219e-06, + "loss": 0.013, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.09933894127607346, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0156, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.09758161753416061, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0145, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.1307433545589447, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0139, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.1336813122034073, + "learning_rate": 9.694397367364635e-06, + "loss": 0.015, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.08841264992952347, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0162, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09142819046974182, + "learning_rate": 9.68666525627968e-06, + "loss": 0.0132, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11312894523143768, + "learning_rate": 9.682799805859474e-06, + "loss": 0.0129, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.14984415471553802, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0147, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.104182668030262, + "learning_rate": 9.675070116731382e-06, + "loss": 0.013, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.10078790783882141, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0131, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.08538635075092316, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0135, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.11059115827083588, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0134, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.10521579533815384, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0131, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.1130848377943039, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0137, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.09023619443178177, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0125, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.11787812411785126, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0136, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.14386765658855438, + "learning_rate": 9.6441675594234e-06, + "loss": 0.0153, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.10471580922603607, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0128, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.1340251863002777, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0139, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.1131536215543747, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0128, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10629499703645706, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0147, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.10359197854995728, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0121, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.11530483514070511, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0136, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.09336108714342117, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0124, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.08179444819688797, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0124, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.12341088801622391, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0125, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.12123409658670425, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0149, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.11406739801168442, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0141, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.07922708988189697, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0114, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.09156256914138794, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0121, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.09705155342817307, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0134, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.10157317668199539, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0157, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.09271441400051117, + "learning_rate": 9.582440688952473e-06, + "loss": 0.0113, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.09375326335430145, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0109, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.07997307926416397, + "learning_rate": 9.57473221124502e-06, + "loss": 0.0137, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.11743918806314468, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0138, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.09150344878435135, + "learning_rate": 9.567025381590669e-06, + "loss": 0.013, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.09968695789575577, + "learning_rate": 9.563172585515032e-06, + "loss": 0.011, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.10516709089279175, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0114, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.09716377407312393, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0132, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.12646295130252838, + "learning_rate": 9.551616675805319e-06, + "loss": 0.0145, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.11036434769630432, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0153, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10279295593500137, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0138, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.08990927040576935, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0132, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.10534624755382538, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0138, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.09507470577955246, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0114, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.101425401866436, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0137, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.11853020638227463, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0121, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.13193756341934204, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0128, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.10581472516059875, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0123, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.12424535304307938, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0139, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.1036083921790123, + "learning_rate": 9.509276927101462e-06, + "loss": 0.0126, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.07990803569555283, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0112, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.098060242831707, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0136, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.11892619729042053, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0126, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.10609585791826248, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0127, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.10905738174915314, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0123, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.0981525182723999, + "learning_rate": 9.486203760067013e-06, + "loss": 0.012, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.16955061256885529, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0119, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.10360468924045563, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0136, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.08537023514509201, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0112, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.1317787766456604, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0128, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10730036348104477, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0151, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09596534818410873, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0117, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.10839905589818954, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0138, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.12511657178401947, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0138, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.1078735962510109, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0127, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.10344097018241882, + "learning_rate": 9.447782011716307e-06, + "loss": 0.0119, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.18708902597427368, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0139, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.09599670022726059, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0125, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.13859155774116516, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0143, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09208878129720688, + "learning_rate": 9.432425102360072e-06, + "loss": 0.0126, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.21019718050956726, + "learning_rate": 9.428586931343613e-06, + "loss": 0.0117, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.09587983787059784, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0122, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.09626271575689316, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0124, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.1051478385925293, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0154, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.10166551172733307, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0119, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.09312118589878082, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0145, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.11722584068775177, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0126, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.0939759686589241, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0124, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.11434069275856018, + "learning_rate": 9.39789681972375e-06, + "loss": 0.012, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12582115828990936, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0137, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.07123387604951859, + "learning_rate": 9.390228540420876e-06, + "loss": 0.0155, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.08926647901535034, + "learning_rate": 9.386395039661539e-06, + "loss": 0.014, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.07371679693460464, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0128, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.1008603498339653, + "learning_rate": 9.378729317382715e-06, + "loss": 0.012, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.10988970100879669, + "learning_rate": 9.37489709644538e-06, + "loss": 0.0136, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.09886831045150757, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0123, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.10959886759519577, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0135, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.12443319708108902, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0161, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10207810997962952, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0146, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.11660879850387573, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0127, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.13432076573371887, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0142, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.11405819654464722, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0135, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.12952250242233276, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0119, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.09084811806678772, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0173, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.12252718955278397, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0099, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1293795108795166, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0144, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11287739127874374, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0164, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.1253224015235901, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0133, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.10046776384115219, + "learning_rate": 9.321290990574668e-06, + "loss": 0.014, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.08746109157800674, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0128, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.08306485414505005, + "learning_rate": 9.313639857014607e-06, + "loss": 0.0115, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.1340319961309433, + "learning_rate": 9.3098149378516e-06, + "loss": 0.0137, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.09587698429822922, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0143, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.14915677905082703, + "learning_rate": 9.302166396212059e-06, + "loss": 0.0143, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.09313366562128067, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0115, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.12605343759059906, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0146, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.12807077169418335, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0113, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.1186203733086586, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0119, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.10223432630300522, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0141, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.12712709605693817, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0132, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.12993593513965607, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0162, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.12273230403661728, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0133, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.1391134113073349, + "learning_rate": 9.26776941165236e-06, + "loss": 0.014, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.11695246398448944, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0137, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.1214752346277237, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0122, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.13961699604988098, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0117, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.1402655988931656, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0159, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.10683967918157578, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0146, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.08510883897542953, + "learning_rate": 9.244857665179213e-06, + "loss": 0.013, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.10308418422937393, + "learning_rate": 9.241040568803432e-06, + "loss": 0.0141, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.09736710786819458, + "learning_rate": 9.23722390978261e-06, + "loss": 0.0139, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.13270588219165802, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0118, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.08694430440664291, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0132, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.07487893849611282, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0123, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.1315941959619522, + "learning_rate": 9.221961653045449e-06, + "loss": 0.0115, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10982084274291992, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0146, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.09258668124675751, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.11609145998954773, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0143, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.1349419355392456, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0139, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.08282078802585602, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0108, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.12237592041492462, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0149, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.07864762842655182, + "learning_rate": 9.195269608854445e-06, + "loss": 0.012, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.1032051146030426, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0112, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.1346377581357956, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0132, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.08922939747571945, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0118, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.09849593788385391, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0141, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.07525688409805298, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0126, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.1287003457546234, + "learning_rate": 9.172407902473605e-06, + "loss": 0.0117, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.11016935855150223, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0123, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.09082796424627304, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0119, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.12462439388036728, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0133, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.10698465257883072, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0126, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.07600131630897522, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0133, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.09345563501119614, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0144, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.1186055988073349, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0151, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.12372780591249466, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0133, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.116783007979393, + "learning_rate": 9.138145235699377e-06, + "loss": 0.0139, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.08260510116815567, + "learning_rate": 9.134340494716546e-06, + "loss": 0.012, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.1815570890903473, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0125, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.12819825112819672, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0133, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.09238126873970032, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0115, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.09430833160877228, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0119, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.7733426094055176, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0128, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.10260160267353058, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0126, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.11146093159914017, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0136, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.11368793249130249, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0141, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.08438809961080551, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0134, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.11850261688232422, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.13420401513576508, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0128, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.09115619212388992, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0127, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.09974579513072968, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0141, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.10242673754692078, + "learning_rate": 9.081121025468816e-06, + "loss": 0.0133, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.08850608766078949, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0113, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.08285540342330933, + "learning_rate": 9.073525417675752e-06, + "loss": 0.0122, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.08347760140895844, + "learning_rate": 9.069728288749965e-06, + "loss": 0.011, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.12695780396461487, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0134, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.10132371634244919, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0143, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.07370374351739883, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0112, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.0993623211979866, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0138, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.07118473947048187, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0134, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.10804349929094315, + "learning_rate": 9.046954982950699e-06, + "loss": 0.014, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.12209299206733704, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0127, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.0896674394607544, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0134, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.10866492241621017, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0145, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.06327113509178162, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0125, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.1044771745800972, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0138, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10678671300411224, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0119, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.10146090388298035, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0122, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.11703896522521973, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0126, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.08116868138313293, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0117, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.16623066365718842, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0135, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.07943270355463028, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0128, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.17520980536937714, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0141, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.10311958938837051, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0124, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09508595615625381, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0139, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.11690590530633926, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0139, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.11585275083780289, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0119, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.08755052834749222, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0152, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.15034662187099457, + "learning_rate": 8.978732966762205e-06, + "loss": 0.013, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.11793015897274017, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0147, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.14666646718978882, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0159, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1565992683172226, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0152, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.11420148611068726, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0134, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.0945156067609787, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0105, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.0971534475684166, + "learning_rate": 8.95602513567626e-06, + "loss": 0.012, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.12953956425189972, + "learning_rate": 8.952242101985645e-06, + "loss": 0.0144, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.11144538223743439, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0138, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.11536341160535812, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0118, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.08281587064266205, + "learning_rate": 8.940895757786624e-06, + "loss": 0.0091, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.10917636007070541, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0116, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.11539111286401749, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0142, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.09897308051586151, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0141, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.12034150213003159, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0142, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.10506895184516907, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0145, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10428126901388168, + "learning_rate": 8.918215501162718e-06, + "loss": 0.013, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.0966438427567482, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0126, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.09688732028007507, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0136, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10319554805755615, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0133, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11589661985635757, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0136, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.08819710463285446, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0145, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.09449224174022675, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0141, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.11158958822488785, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0124, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.10327963531017303, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0138, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.070926733314991, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0143, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.08468566834926605, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0131, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11933926492929459, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0128, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10370025038719177, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0138, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.13985954225063324, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0144, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.08677119761705399, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0141, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.1023884266614914, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0113, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.14462946355342865, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0132, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.12994524836540222, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0144, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.08624777942895889, + "learning_rate": 8.850274743290876e-06, + "loss": 0.013, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.10459248721599579, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0117, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.12475694715976715, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0131, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09973360598087311, + "learning_rate": 8.838965950000907e-06, + "loss": 0.0142, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.10204580426216125, + "learning_rate": 8.835197287635673e-06, + "loss": 0.0136, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.08137936890125275, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0111, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.09507330507040024, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0128, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.11097851395606995, + "learning_rate": 8.823894110741103e-06, + "loss": 0.0145, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.12435153126716614, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0122, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.1186857596039772, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0123, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10177282989025116, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0119, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.08287174254655838, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0129, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.12975555658340454, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0138, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.07198113948106766, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0116, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.12089285254478455, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0126, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.09319751709699631, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0123, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.11510422080755234, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0147, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.1134580671787262, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0119, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12099597603082657, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0116, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.11170639842748642, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0111, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.0979531928896904, + "learning_rate": 8.774962556249487e-06, + "loss": 0.0137, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.0814109742641449, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0143, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.09910739213228226, + "learning_rate": 8.767441700331522e-06, + "loss": 0.0129, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.10014151781797409, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0125, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.14257559180259705, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.08456768840551376, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0138, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.19197505712509155, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0156, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.10880734771490097, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0143, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.12371200323104858, + "learning_rate": 8.744890501433196e-06, + "loss": 0.0117, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.08023890107870102, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0118, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.10231798142194748, + "learning_rate": 8.737377232363901e-06, + "loss": 0.0107, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.06945646554231644, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0131, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.09935662895441055, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0122, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.15378066897392273, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0142, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.08416267484426498, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0127, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.0862744078040123, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0132, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.11481238156557083, + "learning_rate": 8.71484884879911e-06, + "loss": 0.013, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.09441596269607544, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0126, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08091022074222565, + "learning_rate": 8.707343203095797e-06, + "loss": 0.0126, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.1480294167995453, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0121, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.12529131770133972, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0132, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.12804464995861053, + "learning_rate": 8.696088319396272e-06, + "loss": 0.013, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.10956594347953796, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0129, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.11884522438049316, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0127, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.09215931594371796, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0122, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.1411169320344925, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0139, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.07872501015663147, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0139, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.083979532122612, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0102, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.12711060047149658, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0148, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.10543761402368546, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0131, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.13125181198120117, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0142, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.13816513121128082, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0122, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.09294828027486801, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0126, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.10944795608520508, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0142, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.11008825153112411, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0122, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.10732167959213257, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0138, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.10374412685632706, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0136, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.1021609753370285, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0138, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.10304085910320282, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0125, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.08420020341873169, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0119, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.11911525577306747, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0126, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.07106399536132812, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0117, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.08642835915088654, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0125, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1176094189286232, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0146, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08892391622066498, + "learning_rate": 8.60994458849601e-06, + "loss": 0.012, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.12898282706737518, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0122, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.12732256948947906, + "learning_rate": 8.602465911610909e-06, + "loss": 0.0164, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.1723143309354782, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0146, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.10019178688526154, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0122, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.15125800669193268, + "learning_rate": 8.591251540871672e-06, + "loss": 0.0138, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.08618403971195221, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0106, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.12111509591341019, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0125, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.11654779314994812, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0133, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.12341999262571335, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0136, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.09898588061332703, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0121, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.09825480729341507, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0119, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.09357906877994537, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0125, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13970960676670074, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0131, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.14642572402954102, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0117, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.20111511647701263, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0122, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.12268098443746567, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0127, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.11043815314769745, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0109, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.12570181488990784, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0147, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.1037122830748558, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0134, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.07811714708805084, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0121, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.07688374072313309, + "learning_rate": 8.53151572917108e-06, + "loss": 0.013, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.12285313010215759, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0135, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11350435763597488, + "learning_rate": 8.524057571200438e-06, + "loss": 0.013, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.1195593997836113, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.12681595981121063, + "learning_rate": 8.516601380450119e-06, + "loss": 0.0116, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.10786957293748856, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0117, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.12382174283266068, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0123, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08464337885379791, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0139, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.11292237043380737, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0142, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.13033312559127808, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0125, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.13949616253376007, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0153, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.09846726804971695, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0115, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.12679991126060486, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0149, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.10708379000425339, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0145, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.0848502367734909, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0135, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.13396726548671722, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0142, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13248398900032043, + "learning_rate": 8.471905674372712e-06, + "loss": 0.014, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.10274169594049454, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0139, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.08507993817329407, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0112, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.10394402593374252, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0146, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.10293199867010117, + "learning_rate": 8.457022946052628e-06, + "loss": 0.015, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.16253890097141266, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0128, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.07673029601573944, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0125, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.1329779177904129, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0131, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.0754442885518074, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0125, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.09978003799915314, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0124, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.12813764810562134, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0144, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.08480092883110046, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0124, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.10068657249212265, + "learning_rate": 8.42728135848896e-06, + "loss": 0.012, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.08405598253011703, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0122, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.13025622069835663, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0117, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.11026641726493835, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0127, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.10400493443012238, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0138, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.10374867916107178, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0138, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.128962442278862, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0124, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.07732204347848892, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0124, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.14872954785823822, + "learning_rate": 8.39757171680725e-06, + "loss": 0.013, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.13792090117931366, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0175, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.07409467548131943, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0123, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.11308597028255463, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0149, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.1202300488948822, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0139, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.08481356501579285, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0124, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.07334130257368088, + "learning_rate": 8.375310536902511e-06, + "loss": 0.014, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.09884555637836456, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0137, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.16476134955883026, + "learning_rate": 8.36789416540753e-06, + "loss": 0.0142, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.12900881469249725, + "learning_rate": 8.364186735016254e-06, + "loss": 0.0138, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.10485891997814178, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0146, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.13036216795444489, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0111, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.13155606389045715, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0146, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.11586495488882065, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0135, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.1043778583407402, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0137, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.12916314601898193, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.1257125288248062, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0117, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.20916301012039185, + "learning_rate": 8.334545457592674e-06, + "loss": 0.014, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.13241995871067047, + "learning_rate": 8.330842572848802e-06, + "loss": 0.011, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.1633574515581131, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0141, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.08672885596752167, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0141, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.09803608059883118, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0143, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.13284195959568024, + "learning_rate": 8.316036101469748e-06, + "loss": 0.011, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10581238567829132, + "learning_rate": 8.312335751929666e-06, + "loss": 0.0144, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.11561308056116104, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0111, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.09897869825363159, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0116, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.129640132188797, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0142, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.08553853631019592, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.09117261320352554, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0134, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.1338709443807602, + "learning_rate": 8.290144330085352e-06, + "loss": 0.013, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.11192607134580612, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0122, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.0983581617474556, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0149, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.10853386670351028, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0145, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.09507740288972855, + "learning_rate": 8.275360236495652e-06, + "loss": 0.011, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.08634619414806366, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0146, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.12652689218521118, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0145, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.13687895238399506, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0115, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.11267609149217606, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0139, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.0999043732881546, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0125, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.09523805975914001, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0137, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09847134351730347, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0153, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.0827624723315239, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0093, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.09830755740404129, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0119, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.14614145457744598, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0149, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.1254947930574417, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0129, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.11176569014787674, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0147, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.06597425043582916, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0132, + "step": 80000 + }, + { + "epoch": 23.279022403258654, + "grad_norm": 0.19944901764392853, + "learning_rate": 8.223680373918572e-06, + "loss": 0.0147, + "step": 80010 + }, + { + "epoch": 23.2819319173698, + "grad_norm": 0.09820935130119324, + "learning_rate": 8.219992805682155e-06, + "loss": 0.013, + "step": 80020 + }, + { + "epoch": 23.28484143148094, + "grad_norm": 0.07166675478219986, + "learning_rate": 8.216305752342533e-06, + "loss": 0.0135, + "step": 80030 + }, + { + "epoch": 23.287750945592087, + "grad_norm": 0.08634001016616821, + "learning_rate": 8.21261921417972e-06, + "loss": 0.0128, + "step": 80040 + }, + { + "epoch": 23.29066045970323, + "grad_norm": 0.1467038244009018, + "learning_rate": 8.20893319147369e-06, + "loss": 0.0131, + "step": 80050 + }, + { + "epoch": 23.293569973814375, + "grad_norm": 0.0938076600432396, + "learning_rate": 8.205247684504364e-06, + "loss": 0.0122, + "step": 80060 + }, + { + "epoch": 23.296479487925517, + "grad_norm": 0.06792404502630234, + "learning_rate": 8.20156269355164e-06, + "loss": 0.012, + "step": 80070 + }, + { + "epoch": 23.29938900203666, + "grad_norm": 0.06738271564245224, + "learning_rate": 8.197878218895356e-06, + "loss": 0.0114, + "step": 80080 + }, + { + "epoch": 23.302298516147804, + "grad_norm": 0.16149401664733887, + "learning_rate": 8.194194260815338e-06, + "loss": 0.0121, + "step": 80090 + }, + { + "epoch": 23.305208030258946, + "grad_norm": 0.11988461762666702, + "learning_rate": 8.190510819591348e-06, + "loss": 0.0121, + "step": 80100 + }, + { + "epoch": 23.30811754437009, + "grad_norm": 0.09137140959501266, + "learning_rate": 8.186827895503117e-06, + "loss": 0.0125, + "step": 80110 + }, + { + "epoch": 23.311027058481233, + "grad_norm": 0.0880751833319664, + "learning_rate": 8.183145488830355e-06, + "loss": 0.0116, + "step": 80120 + }, + { + "epoch": 23.313936572592375, + "grad_norm": 0.09596525132656097, + "learning_rate": 8.179463599852705e-06, + "loss": 0.0117, + "step": 80130 + }, + { + "epoch": 23.31684608670352, + "grad_norm": 0.13169355690479279, + "learning_rate": 8.175782228849769e-06, + "loss": 0.0121, + "step": 80140 + }, + { + "epoch": 23.319755600814663, + "grad_norm": 0.1044575497508049, + "learning_rate": 8.17210137610115e-06, + "loss": 0.0132, + "step": 80150 + }, + { + "epoch": 23.322665114925808, + "grad_norm": 0.13589617609977722, + "learning_rate": 8.168421041886373e-06, + "loss": 0.0118, + "step": 80160 + }, + { + "epoch": 23.32557462903695, + "grad_norm": 0.10503333061933517, + "learning_rate": 8.164741226484925e-06, + "loss": 0.0108, + "step": 80170 + }, + { + "epoch": 23.328484143148096, + "grad_norm": 0.09190402179956436, + "learning_rate": 8.161061930176272e-06, + "loss": 0.0129, + "step": 80180 + }, + { + "epoch": 23.331393657259238, + "grad_norm": 0.104253388941288, + "learning_rate": 8.15738315323984e-06, + "loss": 0.0114, + "step": 80190 + }, + { + "epoch": 23.33430317137038, + "grad_norm": 0.10191919654607773, + "learning_rate": 8.153704895955e-06, + "loss": 0.0128, + "step": 80200 + }, + { + "epoch": 23.337212685481525, + "grad_norm": 0.0824175626039505, + "learning_rate": 8.150027158601084e-06, + "loss": 0.0126, + "step": 80210 + }, + { + "epoch": 23.340122199592667, + "grad_norm": 0.08966456353664398, + "learning_rate": 8.146349941457401e-06, + "loss": 0.014, + "step": 80220 + }, + { + "epoch": 23.343031713703812, + "grad_norm": 0.0856788232922554, + "learning_rate": 8.142673244803215e-06, + "loss": 0.01, + "step": 80230 + }, + { + "epoch": 23.345941227814954, + "grad_norm": 0.1027766764163971, + "learning_rate": 8.138997068917736e-06, + "loss": 0.0134, + "step": 80240 + }, + { + "epoch": 23.3488507419261, + "grad_norm": 0.11636431515216827, + "learning_rate": 8.13532141408015e-06, + "loss": 0.0126, + "step": 80250 + }, + { + "epoch": 23.35176025603724, + "grad_norm": 0.09742997586727142, + "learning_rate": 8.131646280569605e-06, + "loss": 0.0117, + "step": 80260 + }, + { + "epoch": 23.354669770148384, + "grad_norm": 0.08142542093992233, + "learning_rate": 8.1279716686652e-06, + "loss": 0.0117, + "step": 80270 + }, + { + "epoch": 23.35757928425953, + "grad_norm": 0.09734860807657242, + "learning_rate": 8.124297578645989e-06, + "loss": 0.0115, + "step": 80280 + }, + { + "epoch": 23.36048879837067, + "grad_norm": 0.10368160158395767, + "learning_rate": 8.120624010791002e-06, + "loss": 0.0142, + "step": 80290 + }, + { + "epoch": 23.363398312481817, + "grad_norm": 0.1619454324245453, + "learning_rate": 8.116950965379225e-06, + "loss": 0.0101, + "step": 80300 + }, + { + "epoch": 23.36630782659296, + "grad_norm": 0.09248818457126617, + "learning_rate": 8.113278442689593e-06, + "loss": 0.0112, + "step": 80310 + }, + { + "epoch": 23.369217340704104, + "grad_norm": 0.11162950843572617, + "learning_rate": 8.109606443001019e-06, + "loss": 0.0126, + "step": 80320 + }, + { + "epoch": 23.372126854815246, + "grad_norm": 0.1364763081073761, + "learning_rate": 8.10593496659236e-06, + "loss": 0.014, + "step": 80330 + }, + { + "epoch": 23.375036368926388, + "grad_norm": 0.113436259329319, + "learning_rate": 8.102264013742442e-06, + "loss": 0.0125, + "step": 80340 + }, + { + "epoch": 23.377945883037533, + "grad_norm": 0.11578390002250671, + "learning_rate": 8.098593584730057e-06, + "loss": 0.0118, + "step": 80350 + }, + { + "epoch": 23.380855397148675, + "grad_norm": 0.11272793263196945, + "learning_rate": 8.09492367983394e-06, + "loss": 0.0135, + "step": 80360 + }, + { + "epoch": 23.38376491125982, + "grad_norm": 0.09545538574457169, + "learning_rate": 8.091254299332803e-06, + "loss": 0.0146, + "step": 80370 + }, + { + "epoch": 23.386674425370963, + "grad_norm": 0.10459403693675995, + "learning_rate": 8.087585443505305e-06, + "loss": 0.013, + "step": 80380 + }, + { + "epoch": 23.389583939482108, + "grad_norm": 0.1124049723148346, + "learning_rate": 8.083917112630078e-06, + "loss": 0.0129, + "step": 80390 + }, + { + "epoch": 23.39249345359325, + "grad_norm": 0.34668007493019104, + "learning_rate": 8.0802493069857e-06, + "loss": 0.012, + "step": 80400 + }, + { + "epoch": 23.395402967704392, + "grad_norm": 0.08212853223085403, + "learning_rate": 8.07658202685072e-06, + "loss": 0.0152, + "step": 80410 + }, + { + "epoch": 23.398312481815537, + "grad_norm": 0.10383888334035873, + "learning_rate": 8.072915272503651e-06, + "loss": 0.0117, + "step": 80420 + }, + { + "epoch": 23.40122199592668, + "grad_norm": 0.12535853683948517, + "learning_rate": 8.069249044222951e-06, + "loss": 0.0152, + "step": 80430 + }, + { + "epoch": 23.404131510037825, + "grad_norm": 0.1187693327665329, + "learning_rate": 8.065583342287044e-06, + "loss": 0.0137, + "step": 80440 + }, + { + "epoch": 23.407041024148967, + "grad_norm": 0.1197381466627121, + "learning_rate": 8.061918166974316e-06, + "loss": 0.0124, + "step": 80450 + }, + { + "epoch": 23.409950538260112, + "grad_norm": 0.09074021875858307, + "learning_rate": 8.058253518563122e-06, + "loss": 0.0145, + "step": 80460 + }, + { + "epoch": 23.412860052371254, + "grad_norm": 0.10440906882286072, + "learning_rate": 8.054589397331757e-06, + "loss": 0.0127, + "step": 80470 + }, + { + "epoch": 23.415769566482396, + "grad_norm": 0.09512275457382202, + "learning_rate": 8.050925803558491e-06, + "loss": 0.0142, + "step": 80480 + }, + { + "epoch": 23.41867908059354, + "grad_norm": 0.06672519445419312, + "learning_rate": 8.047262737521556e-06, + "loss": 0.0119, + "step": 80490 + }, + { + "epoch": 23.421588594704684, + "grad_norm": 0.0820726528763771, + "learning_rate": 8.043600199499131e-06, + "loss": 0.0118, + "step": 80500 + }, + { + "epoch": 23.42449810881583, + "grad_norm": 0.10173840820789337, + "learning_rate": 8.03993818976936e-06, + "loss": 0.0133, + "step": 80510 + }, + { + "epoch": 23.42740762292697, + "grad_norm": 0.06551963835954666, + "learning_rate": 8.03627670861035e-06, + "loss": 0.0111, + "step": 80520 + }, + { + "epoch": 23.430317137038113, + "grad_norm": 0.08851084858179092, + "learning_rate": 8.03261575630017e-06, + "loss": 0.011, + "step": 80530 + }, + { + "epoch": 23.43322665114926, + "grad_norm": 0.11487096548080444, + "learning_rate": 8.028955333116846e-06, + "loss": 0.011, + "step": 80540 + }, + { + "epoch": 23.4361361652604, + "grad_norm": 0.08430162072181702, + "learning_rate": 8.025295439338347e-06, + "loss": 0.0119, + "step": 80550 + }, + { + "epoch": 23.439045679371546, + "grad_norm": 0.06152086704969406, + "learning_rate": 8.021636075242641e-06, + "loss": 0.0122, + "step": 80560 + }, + { + "epoch": 23.441955193482688, + "grad_norm": 0.11052606254816055, + "learning_rate": 8.017977241107622e-06, + "loss": 0.0136, + "step": 80570 + }, + { + "epoch": 23.444864707593833, + "grad_norm": 0.1052686870098114, + "learning_rate": 8.01431893721115e-06, + "loss": 0.012, + "step": 80580 + }, + { + "epoch": 23.447774221704975, + "grad_norm": 0.10118360817432404, + "learning_rate": 8.01066116383105e-06, + "loss": 0.014, + "step": 80590 + }, + { + "epoch": 23.450683735816117, + "grad_norm": 0.10555325448513031, + "learning_rate": 8.00700392124512e-06, + "loss": 0.013, + "step": 80600 + }, + { + "epoch": 23.453593249927263, + "grad_norm": 0.06563732028007507, + "learning_rate": 8.003347209731084e-06, + "loss": 0.0122, + "step": 80610 + }, + { + "epoch": 23.456502764038405, + "grad_norm": 0.1160956397652626, + "learning_rate": 7.999691029566658e-06, + "loss": 0.0125, + "step": 80620 + }, + { + "epoch": 23.45941227814955, + "grad_norm": 0.08643034845590591, + "learning_rate": 7.996035381029494e-06, + "loss": 0.0135, + "step": 80630 + }, + { + "epoch": 23.462321792260692, + "grad_norm": 0.09050534665584564, + "learning_rate": 7.99238026439723e-06, + "loss": 0.0124, + "step": 80640 + }, + { + "epoch": 23.465231306371837, + "grad_norm": 0.07935993373394012, + "learning_rate": 7.988725679947432e-06, + "loss": 0.0129, + "step": 80650 + }, + { + "epoch": 23.46814082048298, + "grad_norm": 0.11728908121585846, + "learning_rate": 7.985071627957645e-06, + "loss": 0.0122, + "step": 80660 + }, + { + "epoch": 23.47105033459412, + "grad_norm": 0.18306072056293488, + "learning_rate": 7.981418108705381e-06, + "loss": 0.0131, + "step": 80670 + }, + { + "epoch": 23.473959848705267, + "grad_norm": 0.10635816305875778, + "learning_rate": 7.977765122468089e-06, + "loss": 0.0119, + "step": 80680 + }, + { + "epoch": 23.47686936281641, + "grad_norm": 0.13399194180965424, + "learning_rate": 7.974112669523197e-06, + "loss": 0.0164, + "step": 80690 + }, + { + "epoch": 23.479778876927554, + "grad_norm": 0.12431732565164566, + "learning_rate": 7.970460750148077e-06, + "loss": 0.0112, + "step": 80700 + }, + { + "epoch": 23.482688391038696, + "grad_norm": 0.12431129813194275, + "learning_rate": 7.96680936462007e-06, + "loss": 0.0143, + "step": 80710 + }, + { + "epoch": 23.48559790514984, + "grad_norm": 0.09648348391056061, + "learning_rate": 7.963158513216482e-06, + "loss": 0.0105, + "step": 80720 + }, + { + "epoch": 23.488507419260984, + "grad_norm": 0.11739730834960938, + "learning_rate": 7.959508196214568e-06, + "loss": 0.0131, + "step": 80730 + }, + { + "epoch": 23.491416933372125, + "grad_norm": 0.109571672976017, + "learning_rate": 7.955858413891536e-06, + "loss": 0.0124, + "step": 80740 + }, + { + "epoch": 23.49432644748327, + "grad_norm": 0.09891501069068909, + "learning_rate": 7.952209166524571e-06, + "loss": 0.0123, + "step": 80750 + }, + { + "epoch": 23.497235961594413, + "grad_norm": 0.11239849030971527, + "learning_rate": 7.948560454390814e-06, + "loss": 0.0114, + "step": 80760 + }, + { + "epoch": 23.50014547570556, + "grad_norm": 0.08035823702812195, + "learning_rate": 7.94491227776735e-06, + "loss": 0.0106, + "step": 80770 + }, + { + "epoch": 23.5030549898167, + "grad_norm": 0.1137743890285492, + "learning_rate": 7.941264636931237e-06, + "loss": 0.0125, + "step": 80780 + }, + { + "epoch": 23.505964503927846, + "grad_norm": 0.11350683122873306, + "learning_rate": 7.9376175321595e-06, + "loss": 0.0119, + "step": 80790 + }, + { + "epoch": 23.508874018038988, + "grad_norm": 0.15854588150978088, + "learning_rate": 7.933970963729102e-06, + "loss": 0.0111, + "step": 80800 + }, + { + "epoch": 23.51178353215013, + "grad_norm": 0.09225080907344818, + "learning_rate": 7.93032493191697e-06, + "loss": 0.0145, + "step": 80810 + }, + { + "epoch": 23.514693046261275, + "grad_norm": 0.07517179846763611, + "learning_rate": 7.926679437000006e-06, + "loss": 0.0129, + "step": 80820 + }, + { + "epoch": 23.517602560372417, + "grad_norm": 0.09395122528076172, + "learning_rate": 7.923034479255065e-06, + "loss": 0.014, + "step": 80830 + }, + { + "epoch": 23.520512074483563, + "grad_norm": 0.0953364223241806, + "learning_rate": 7.91939005895895e-06, + "loss": 0.0132, + "step": 80840 + }, + { + "epoch": 23.523421588594704, + "grad_norm": 0.1388244777917862, + "learning_rate": 7.91574617638842e-06, + "loss": 0.0142, + "step": 80850 + }, + { + "epoch": 23.526331102705846, + "grad_norm": 0.10685098171234131, + "learning_rate": 7.912102831820231e-06, + "loss": 0.0143, + "step": 80860 + }, + { + "epoch": 23.529240616816992, + "grad_norm": 0.09484872967004776, + "learning_rate": 7.908460025531051e-06, + "loss": 0.0138, + "step": 80870 + }, + { + "epoch": 23.532150130928134, + "grad_norm": 0.0918048843741417, + "learning_rate": 7.90481775779753e-06, + "loss": 0.0136, + "step": 80880 + }, + { + "epoch": 23.53505964503928, + "grad_norm": 0.15061822533607483, + "learning_rate": 7.901176028896274e-06, + "loss": 0.0149, + "step": 80890 + }, + { + "epoch": 23.53796915915042, + "grad_norm": 0.0951741561293602, + "learning_rate": 7.897534839103856e-06, + "loss": 0.0118, + "step": 80900 + }, + { + "epoch": 23.540878673261567, + "grad_norm": 0.08154623210430145, + "learning_rate": 7.893894188696797e-06, + "loss": 0.0144, + "step": 80910 + }, + { + "epoch": 23.54378818737271, + "grad_norm": 0.10864793509244919, + "learning_rate": 7.89025407795157e-06, + "loss": 0.0132, + "step": 80920 + }, + { + "epoch": 23.54669770148385, + "grad_norm": 0.12371568381786346, + "learning_rate": 7.886614507144628e-06, + "loss": 0.0128, + "step": 80930 + }, + { + "epoch": 23.549607215594996, + "grad_norm": 0.13036499917507172, + "learning_rate": 7.882975476552372e-06, + "loss": 0.0135, + "step": 80940 + }, + { + "epoch": 23.552516729706138, + "grad_norm": 0.10305178165435791, + "learning_rate": 7.87933698645116e-06, + "loss": 0.0104, + "step": 80950 + }, + { + "epoch": 23.555426243817283, + "grad_norm": 0.09468064457178116, + "learning_rate": 7.875699037117306e-06, + "loss": 0.0131, + "step": 80960 + }, + { + "epoch": 23.558335757928425, + "grad_norm": 0.12249180674552917, + "learning_rate": 7.872061628827103e-06, + "loss": 0.0132, + "step": 80970 + }, + { + "epoch": 23.56124527203957, + "grad_norm": 0.0880570113658905, + "learning_rate": 7.86842476185677e-06, + "loss": 0.0103, + "step": 80980 + }, + { + "epoch": 23.564154786150713, + "grad_norm": 0.12612831592559814, + "learning_rate": 7.864788436482522e-06, + "loss": 0.0123, + "step": 80990 + }, + { + "epoch": 23.567064300261855, + "grad_norm": 0.12270201742649078, + "learning_rate": 7.861152652980497e-06, + "loss": 0.0121, + "step": 81000 + }, + { + "epoch": 23.569973814373, + "grad_norm": 0.10333582013845444, + "learning_rate": 7.85751741162682e-06, + "loss": 0.0098, + "step": 81010 + }, + { + "epoch": 23.572883328484142, + "grad_norm": 0.15205809473991394, + "learning_rate": 7.853882712697554e-06, + "loss": 0.0144, + "step": 81020 + }, + { + "epoch": 23.575792842595288, + "grad_norm": 0.09015808999538422, + "learning_rate": 7.850248556468741e-06, + "loss": 0.0109, + "step": 81030 + }, + { + "epoch": 23.57870235670643, + "grad_norm": 0.11366626620292664, + "learning_rate": 7.846614943216361e-06, + "loss": 0.0121, + "step": 81040 + }, + { + "epoch": 23.581611870817575, + "grad_norm": 0.08618906885385513, + "learning_rate": 7.842981873216368e-06, + "loss": 0.0137, + "step": 81050 + }, + { + "epoch": 23.584521384928717, + "grad_norm": 0.11496937274932861, + "learning_rate": 7.839349346744675e-06, + "loss": 0.0144, + "step": 81060 + }, + { + "epoch": 23.58743089903986, + "grad_norm": 0.11915834993124008, + "learning_rate": 7.835717364077138e-06, + "loss": 0.014, + "step": 81070 + }, + { + "epoch": 23.590340413151004, + "grad_norm": 0.12172401696443558, + "learning_rate": 7.832085925489584e-06, + "loss": 0.0125, + "step": 81080 + }, + { + "epoch": 23.593249927262146, + "grad_norm": 0.10628852248191833, + "learning_rate": 7.82845503125781e-06, + "loss": 0.0135, + "step": 81090 + }, + { + "epoch": 23.596159441373292, + "grad_norm": 0.11503509432077408, + "learning_rate": 7.824824681657544e-06, + "loss": 0.0128, + "step": 81100 + }, + { + "epoch": 23.599068955484434, + "grad_norm": 0.11923188716173172, + "learning_rate": 7.821194876964489e-06, + "loss": 0.0132, + "step": 81110 + }, + { + "epoch": 23.60197846959558, + "grad_norm": 0.09143807739019394, + "learning_rate": 7.817565617454305e-06, + "loss": 0.0131, + "step": 81120 + }, + { + "epoch": 23.60488798370672, + "grad_norm": 0.07396473735570908, + "learning_rate": 7.813936903402619e-06, + "loss": 0.0122, + "step": 81130 + }, + { + "epoch": 23.607797497817863, + "grad_norm": 0.11674617975950241, + "learning_rate": 7.810308735085001e-06, + "loss": 0.0127, + "step": 81140 + }, + { + "epoch": 23.61070701192901, + "grad_norm": 0.12394711375236511, + "learning_rate": 7.806681112776976e-06, + "loss": 0.0127, + "step": 81150 + }, + { + "epoch": 23.61361652604015, + "grad_norm": 0.14487721025943756, + "learning_rate": 7.803054036754058e-06, + "loss": 0.0124, + "step": 81160 + }, + { + "epoch": 23.616526040151296, + "grad_norm": 0.1078348383307457, + "learning_rate": 7.799427507291694e-06, + "loss": 0.0143, + "step": 81170 + }, + { + "epoch": 23.619435554262438, + "grad_norm": 0.10270193219184875, + "learning_rate": 7.795801524665284e-06, + "loss": 0.0115, + "step": 81180 + }, + { + "epoch": 23.622345068373583, + "grad_norm": 0.06343881785869598, + "learning_rate": 7.792176089150203e-06, + "loss": 0.0129, + "step": 81190 + }, + { + "epoch": 23.625254582484725, + "grad_norm": 0.08612993359565735, + "learning_rate": 7.788551201021787e-06, + "loss": 0.0116, + "step": 81200 + }, + { + "epoch": 23.628164096595867, + "grad_norm": 0.1368905007839203, + "learning_rate": 7.784926860555315e-06, + "loss": 0.014, + "step": 81210 + }, + { + "epoch": 23.631073610707013, + "grad_norm": 0.11004527658224106, + "learning_rate": 7.781303068026029e-06, + "loss": 0.0118, + "step": 81220 + }, + { + "epoch": 23.633983124818155, + "grad_norm": 0.11310701817274094, + "learning_rate": 7.777679823709134e-06, + "loss": 0.0149, + "step": 81230 + }, + { + "epoch": 23.6368926389293, + "grad_norm": 0.09327834844589233, + "learning_rate": 7.7740571278798e-06, + "loss": 0.0136, + "step": 81240 + }, + { + "epoch": 23.639802153040442, + "grad_norm": 0.15093250572681427, + "learning_rate": 7.770434980813133e-06, + "loss": 0.0133, + "step": 81250 + }, + { + "epoch": 23.642711667151584, + "grad_norm": 0.09899720549583435, + "learning_rate": 7.76681338278422e-06, + "loss": 0.0117, + "step": 81260 + }, + { + "epoch": 23.64562118126273, + "grad_norm": 0.11601489782333374, + "learning_rate": 7.7631923340681e-06, + "loss": 0.0128, + "step": 81270 + }, + { + "epoch": 23.64853069537387, + "grad_norm": 0.09682130813598633, + "learning_rate": 7.759571834939765e-06, + "loss": 0.0121, + "step": 81280 + }, + { + "epoch": 23.651440209485017, + "grad_norm": 0.10942123830318451, + "learning_rate": 7.75595188567416e-06, + "loss": 0.0117, + "step": 81290 + }, + { + "epoch": 23.65434972359616, + "grad_norm": 0.08153918385505676, + "learning_rate": 7.7523324865462e-06, + "loss": 0.0131, + "step": 81300 + }, + { + "epoch": 23.657259237707304, + "grad_norm": 0.08599690347909927, + "learning_rate": 7.748713637830768e-06, + "loss": 0.0115, + "step": 81310 + }, + { + "epoch": 23.660168751818446, + "grad_norm": 0.08310934901237488, + "learning_rate": 7.745095339802672e-06, + "loss": 0.0127, + "step": 81320 + }, + { + "epoch": 23.663078265929588, + "grad_norm": 0.10698913782835007, + "learning_rate": 7.741477592736712e-06, + "loss": 0.0126, + "step": 81330 + }, + { + "epoch": 23.665987780040734, + "grad_norm": 0.14655916392803192, + "learning_rate": 7.737860396907623e-06, + "loss": 0.014, + "step": 81340 + }, + { + "epoch": 23.668897294151876, + "grad_norm": 0.07894794642925262, + "learning_rate": 7.73424375259011e-06, + "loss": 0.0116, + "step": 81350 + }, + { + "epoch": 23.67180680826302, + "grad_norm": 0.09374310821294785, + "learning_rate": 7.730627660058841e-06, + "loss": 0.0101, + "step": 81360 + }, + { + "epoch": 23.674716322374163, + "grad_norm": 0.1619149148464203, + "learning_rate": 7.72701211958842e-06, + "loss": 0.0131, + "step": 81370 + }, + { + "epoch": 23.67762583648531, + "grad_norm": 0.0957368016242981, + "learning_rate": 7.723397131453438e-06, + "loss": 0.0119, + "step": 81380 + }, + { + "epoch": 23.68053535059645, + "grad_norm": 0.10419198870658875, + "learning_rate": 7.719782695928416e-06, + "loss": 0.0126, + "step": 81390 + }, + { + "epoch": 23.683444864707592, + "grad_norm": 0.1281127780675888, + "learning_rate": 7.716168813287858e-06, + "loss": 0.0142, + "step": 81400 + }, + { + "epoch": 23.686354378818738, + "grad_norm": 0.10857229679822922, + "learning_rate": 7.712555483806205e-06, + "loss": 0.0141, + "step": 81410 + }, + { + "epoch": 23.68926389292988, + "grad_norm": 0.10917466878890991, + "learning_rate": 7.70894270775787e-06, + "loss": 0.0127, + "step": 81420 + }, + { + "epoch": 23.692173407041025, + "grad_norm": 0.09018818289041519, + "learning_rate": 7.705330485417224e-06, + "loss": 0.0126, + "step": 81430 + }, + { + "epoch": 23.695082921152167, + "grad_norm": 0.15100347995758057, + "learning_rate": 7.70171881705859e-06, + "loss": 0.0146, + "step": 81440 + }, + { + "epoch": 23.697992435263313, + "grad_norm": 0.08587733656167984, + "learning_rate": 7.698107702956232e-06, + "loss": 0.0129, + "step": 81450 + }, + { + "epoch": 23.700901949374455, + "grad_norm": 0.1320592761039734, + "learning_rate": 7.69449714338442e-06, + "loss": 0.0123, + "step": 81460 + }, + { + "epoch": 23.703811463485597, + "grad_norm": 0.1478150337934494, + "learning_rate": 7.690887138617335e-06, + "loss": 0.0133, + "step": 81470 + }, + { + "epoch": 23.706720977596742, + "grad_norm": 0.11000087857246399, + "learning_rate": 7.687277688929134e-06, + "loss": 0.0113, + "step": 81480 + }, + { + "epoch": 23.709630491707884, + "grad_norm": 0.11071797460317612, + "learning_rate": 7.68366879459393e-06, + "loss": 0.0128, + "step": 81490 + }, + { + "epoch": 23.71254000581903, + "grad_norm": 0.10804122686386108, + "learning_rate": 7.680060455885804e-06, + "loss": 0.0136, + "step": 81500 + }, + { + "epoch": 23.71544951993017, + "grad_norm": 0.11369496583938599, + "learning_rate": 7.676452673078779e-06, + "loss": 0.0118, + "step": 81510 + }, + { + "epoch": 23.718359034041313, + "grad_norm": 0.09902388602495193, + "learning_rate": 7.672845446446835e-06, + "loss": 0.0118, + "step": 81520 + }, + { + "epoch": 23.72126854815246, + "grad_norm": 0.1289895921945572, + "learning_rate": 7.669238776263927e-06, + "loss": 0.0134, + "step": 81530 + }, + { + "epoch": 23.7241780622636, + "grad_norm": 0.13489077985286713, + "learning_rate": 7.665632662803961e-06, + "loss": 0.0115, + "step": 81540 + }, + { + "epoch": 23.727087576374746, + "grad_norm": 0.06097550690174103, + "learning_rate": 7.662027106340785e-06, + "loss": 0.0136, + "step": 81550 + }, + { + "epoch": 23.729997090485888, + "grad_norm": 0.09678100794553757, + "learning_rate": 7.658422107148222e-06, + "loss": 0.0098, + "step": 81560 + }, + { + "epoch": 23.732906604597034, + "grad_norm": 0.09136251360177994, + "learning_rate": 7.654817665500058e-06, + "loss": 0.013, + "step": 81570 + }, + { + "epoch": 23.735816118708176, + "grad_norm": 0.09218855947256088, + "learning_rate": 7.651213781670016e-06, + "loss": 0.0127, + "step": 81580 + }, + { + "epoch": 23.738725632819317, + "grad_norm": 0.07712986320257187, + "learning_rate": 7.647610455931784e-06, + "loss": 0.0123, + "step": 81590 + }, + { + "epoch": 23.741635146930463, + "grad_norm": 0.0797911211848259, + "learning_rate": 7.644007688559018e-06, + "loss": 0.0113, + "step": 81600 + }, + { + "epoch": 23.744544661041605, + "grad_norm": 0.09198153764009476, + "learning_rate": 7.64040547982533e-06, + "loss": 0.0126, + "step": 81610 + }, + { + "epoch": 23.74745417515275, + "grad_norm": 0.11342693865299225, + "learning_rate": 7.636803830004267e-06, + "loss": 0.0126, + "step": 81620 + }, + { + "epoch": 23.750363689263892, + "grad_norm": 0.08946854621171951, + "learning_rate": 7.633202739369368e-06, + "loss": 0.0107, + "step": 81630 + }, + { + "epoch": 23.753273203375038, + "grad_norm": 0.10234414041042328, + "learning_rate": 7.6296022081941e-06, + "loss": 0.0114, + "step": 81640 + }, + { + "epoch": 23.75618271748618, + "grad_norm": 0.1319761723279953, + "learning_rate": 7.6260022367519065e-06, + "loss": 0.0124, + "step": 81650 + }, + { + "epoch": 23.75909223159732, + "grad_norm": 0.11398373544216156, + "learning_rate": 7.622402825316175e-06, + "loss": 0.0125, + "step": 81660 + }, + { + "epoch": 23.762001745708467, + "grad_norm": 0.10755864530801773, + "learning_rate": 7.61880397416026e-06, + "loss": 0.0124, + "step": 81670 + }, + { + "epoch": 23.76491125981961, + "grad_norm": 0.12102973461151123, + "learning_rate": 7.61520568355748e-06, + "loss": 0.0115, + "step": 81680 + }, + { + "epoch": 23.767820773930755, + "grad_norm": 0.14041413366794586, + "learning_rate": 7.611607953781084e-06, + "loss": 0.0131, + "step": 81690 + }, + { + "epoch": 23.770730288041896, + "grad_norm": 0.11478327959775925, + "learning_rate": 7.6080107851043116e-06, + "loss": 0.0127, + "step": 81700 + }, + { + "epoch": 23.773639802153042, + "grad_norm": 0.14033357799053192, + "learning_rate": 7.6044141778003314e-06, + "loss": 0.015, + "step": 81710 + }, + { + "epoch": 23.776549316264184, + "grad_norm": 0.11536005139350891, + "learning_rate": 7.600818132142289e-06, + "loss": 0.0141, + "step": 81720 + }, + { + "epoch": 23.779458830375326, + "grad_norm": 0.10734536498785019, + "learning_rate": 7.597222648403283e-06, + "loss": 0.0117, + "step": 81730 + }, + { + "epoch": 23.78236834448647, + "grad_norm": 0.0891343355178833, + "learning_rate": 7.593627726856365e-06, + "loss": 0.0117, + "step": 81740 + }, + { + "epoch": 23.785277858597613, + "grad_norm": 0.09240922331809998, + "learning_rate": 7.5900333677745335e-06, + "loss": 0.0104, + "step": 81750 + }, + { + "epoch": 23.78818737270876, + "grad_norm": 0.10601887851953506, + "learning_rate": 7.586439571430771e-06, + "loss": 0.0137, + "step": 81760 + }, + { + "epoch": 23.7910968868199, + "grad_norm": 0.12168681621551514, + "learning_rate": 7.5828463380980015e-06, + "loss": 0.014, + "step": 81770 + }, + { + "epoch": 23.794006400931046, + "grad_norm": 0.11172056198120117, + "learning_rate": 7.579253668049098e-06, + "loss": 0.0145, + "step": 81780 + }, + { + "epoch": 23.796915915042188, + "grad_norm": 0.1011413112282753, + "learning_rate": 7.575661561556905e-06, + "loss": 0.0116, + "step": 81790 + }, + { + "epoch": 23.79982542915333, + "grad_norm": 0.11835134774446487, + "learning_rate": 7.572070018894225e-06, + "loss": 0.0137, + "step": 81800 + }, + { + "epoch": 23.802734943264475, + "grad_norm": 0.1536318063735962, + "learning_rate": 7.5684790403338096e-06, + "loss": 0.012, + "step": 81810 + }, + { + "epoch": 23.805644457375617, + "grad_norm": 0.0950169786810875, + "learning_rate": 7.564888626148361e-06, + "loss": 0.0118, + "step": 81820 + }, + { + "epoch": 23.808553971486763, + "grad_norm": 0.11085739731788635, + "learning_rate": 7.561298776610552e-06, + "loss": 0.0124, + "step": 81830 + }, + { + "epoch": 23.811463485597905, + "grad_norm": 0.11517464369535446, + "learning_rate": 7.557709491993017e-06, + "loss": 0.0146, + "step": 81840 + }, + { + "epoch": 23.81437299970905, + "grad_norm": 0.14310887455940247, + "learning_rate": 7.55412077256833e-06, + "loss": 0.0141, + "step": 81850 + }, + { + "epoch": 23.817282513820192, + "grad_norm": 0.0991383045911789, + "learning_rate": 7.550532618609022e-06, + "loss": 0.0119, + "step": 81860 + }, + { + "epoch": 23.820192027931334, + "grad_norm": 0.08636868745088577, + "learning_rate": 7.54694503038761e-06, + "loss": 0.0129, + "step": 81870 + }, + { + "epoch": 23.82310154204248, + "grad_norm": 0.11465732753276825, + "learning_rate": 7.543358008176537e-06, + "loss": 0.0117, + "step": 81880 + }, + { + "epoch": 23.82601105615362, + "grad_norm": 0.15294228494167328, + "learning_rate": 7.539771552248207e-06, + "loss": 0.0104, + "step": 81890 + }, + { + "epoch": 23.828920570264767, + "grad_norm": 0.11894837021827698, + "learning_rate": 7.536185662874997e-06, + "loss": 0.0142, + "step": 81900 + }, + { + "epoch": 23.83183008437591, + "grad_norm": 0.11275901645421982, + "learning_rate": 7.532600340329232e-06, + "loss": 0.0119, + "step": 81910 + }, + { + "epoch": 23.834739598487054, + "grad_norm": 0.08029196411371231, + "learning_rate": 7.529015584883188e-06, + "loss": 0.0116, + "step": 81920 + }, + { + "epoch": 23.837649112598196, + "grad_norm": 0.11938098818063736, + "learning_rate": 7.5254313968091095e-06, + "loss": 0.0127, + "step": 81930 + }, + { + "epoch": 23.84055862670934, + "grad_norm": 0.09227802604436874, + "learning_rate": 7.521847776379183e-06, + "loss": 0.0134, + "step": 81940 + }, + { + "epoch": 23.843468140820484, + "grad_norm": 0.10166507214307785, + "learning_rate": 7.518264723865572e-06, + "loss": 0.0124, + "step": 81950 + }, + { + "epoch": 23.846377654931626, + "grad_norm": 0.13323524594306946, + "learning_rate": 7.514682239540374e-06, + "loss": 0.0135, + "step": 81960 + }, + { + "epoch": 23.84928716904277, + "grad_norm": 0.18477675318717957, + "learning_rate": 7.5111003236756625e-06, + "loss": 0.0141, + "step": 81970 + }, + { + "epoch": 23.852196683153913, + "grad_norm": 0.11462473124265671, + "learning_rate": 7.507518976543463e-06, + "loss": 0.0121, + "step": 81980 + }, + { + "epoch": 23.855106197265055, + "grad_norm": 0.14673922955989838, + "learning_rate": 7.503938198415746e-06, + "loss": 0.0131, + "step": 81990 + }, + { + "epoch": 23.8580157113762, + "grad_norm": 0.11723819375038147, + "learning_rate": 7.5003579895644596e-06, + "loss": 0.015, + "step": 82000 + }, + { + "epoch": 23.860925225487343, + "grad_norm": 0.1134917214512825, + "learning_rate": 7.496778350261485e-06, + "loss": 0.0135, + "step": 82010 + }, + { + "epoch": 23.863834739598488, + "grad_norm": 0.12128591537475586, + "learning_rate": 7.493199280778683e-06, + "loss": 0.013, + "step": 82020 + }, + { + "epoch": 23.86674425370963, + "grad_norm": 0.13865070044994354, + "learning_rate": 7.489620781387849e-06, + "loss": 0.0142, + "step": 82030 + }, + { + "epoch": 23.869653767820775, + "grad_norm": 0.0896529033780098, + "learning_rate": 7.486042852360758e-06, + "loss": 0.0109, + "step": 82040 + }, + { + "epoch": 23.872563281931917, + "grad_norm": 0.09412191063165665, + "learning_rate": 7.482465493969121e-06, + "loss": 0.0138, + "step": 82050 + }, + { + "epoch": 23.87547279604306, + "grad_norm": 0.1172008290886879, + "learning_rate": 7.4788887064846155e-06, + "loss": 0.0127, + "step": 82060 + }, + { + "epoch": 23.878382310154205, + "grad_norm": 0.1011468768119812, + "learning_rate": 7.4753124901788845e-06, + "loss": 0.012, + "step": 82070 + }, + { + "epoch": 23.881291824265347, + "grad_norm": 0.10312607884407043, + "learning_rate": 7.471736845323507e-06, + "loss": 0.0135, + "step": 82080 + }, + { + "epoch": 23.884201338376492, + "grad_norm": 0.08789291232824326, + "learning_rate": 7.4681617721900324e-06, + "loss": 0.0141, + "step": 82090 + }, + { + "epoch": 23.887110852487634, + "grad_norm": 0.08846236020326614, + "learning_rate": 7.4645872710499705e-06, + "loss": 0.0098, + "step": 82100 + }, + { + "epoch": 23.89002036659878, + "grad_norm": 0.1339399814605713, + "learning_rate": 7.4610133421747765e-06, + "loss": 0.0142, + "step": 82110 + }, + { + "epoch": 23.89292988070992, + "grad_norm": 0.13560190796852112, + "learning_rate": 7.457439985835862e-06, + "loss": 0.0133, + "step": 82120 + }, + { + "epoch": 23.895839394821063, + "grad_norm": 0.08713413774967194, + "learning_rate": 7.453867202304602e-06, + "loss": 0.0122, + "step": 82130 + }, + { + "epoch": 23.89874890893221, + "grad_norm": 0.10207775980234146, + "learning_rate": 7.450294991852334e-06, + "loss": 0.014, + "step": 82140 + }, + { + "epoch": 23.90165842304335, + "grad_norm": 0.10218358039855957, + "learning_rate": 7.446723354750337e-06, + "loss": 0.0114, + "step": 82150 + }, + { + "epoch": 23.904567937154496, + "grad_norm": 0.11706096678972244, + "learning_rate": 7.443152291269844e-06, + "loss": 0.0118, + "step": 82160 + }, + { + "epoch": 23.90747745126564, + "grad_norm": 0.10717874765396118, + "learning_rate": 7.439581801682075e-06, + "loss": 0.0139, + "step": 82170 + }, + { + "epoch": 23.910386965376784, + "grad_norm": 0.09272971004247665, + "learning_rate": 7.436011886258174e-06, + "loss": 0.0118, + "step": 82180 + }, + { + "epoch": 23.913296479487926, + "grad_norm": 0.08789035677909851, + "learning_rate": 7.4324425452692445e-06, + "loss": 0.0119, + "step": 82190 + }, + { + "epoch": 23.916205993599068, + "grad_norm": 0.10102546960115433, + "learning_rate": 7.428873778986364e-06, + "loss": 0.0127, + "step": 82200 + }, + { + "epoch": 23.919115507710213, + "grad_norm": 0.14371797442436218, + "learning_rate": 7.42530558768056e-06, + "loss": 0.012, + "step": 82210 + }, + { + "epoch": 23.922025021821355, + "grad_norm": 0.09243465960025787, + "learning_rate": 7.4217379716228075e-06, + "loss": 0.0129, + "step": 82220 + }, + { + "epoch": 23.9249345359325, + "grad_norm": 0.13050049543380737, + "learning_rate": 7.418170931084039e-06, + "loss": 0.0116, + "step": 82230 + }, + { + "epoch": 23.927844050043642, + "grad_norm": 0.13714571297168732, + "learning_rate": 7.414604466335152e-06, + "loss": 0.0147, + "step": 82240 + }, + { + "epoch": 23.930753564154784, + "grad_norm": 0.131483793258667, + "learning_rate": 7.411038577647002e-06, + "loss": 0.012, + "step": 82250 + }, + { + "epoch": 23.93366307826593, + "grad_norm": 0.1075965166091919, + "learning_rate": 7.407473265290385e-06, + "loss": 0.014, + "step": 82260 + }, + { + "epoch": 23.936572592377072, + "grad_norm": 0.08791981637477875, + "learning_rate": 7.403908529536066e-06, + "loss": 0.0106, + "step": 82270 + }, + { + "epoch": 23.939482106488217, + "grad_norm": 0.14243891835212708, + "learning_rate": 7.40034437065477e-06, + "loss": 0.0125, + "step": 82280 + }, + { + "epoch": 23.94239162059936, + "grad_norm": 0.11222469806671143, + "learning_rate": 7.396780788917161e-06, + "loss": 0.0126, + "step": 82290 + }, + { + "epoch": 23.945301134710505, + "grad_norm": 0.1121295914053917, + "learning_rate": 7.393217784593878e-06, + "loss": 0.0131, + "step": 82300 + }, + { + "epoch": 23.948210648821647, + "grad_norm": 0.09923163801431656, + "learning_rate": 7.389655357955499e-06, + "loss": 0.0118, + "step": 82310 + }, + { + "epoch": 23.95112016293279, + "grad_norm": 0.14223536849021912, + "learning_rate": 7.386093509272578e-06, + "loss": 0.0127, + "step": 82320 + }, + { + "epoch": 23.954029677043934, + "grad_norm": 0.11418648064136505, + "learning_rate": 7.3825322388156016e-06, + "loss": 0.0128, + "step": 82330 + }, + { + "epoch": 23.956939191155076, + "grad_norm": 0.13791611790657043, + "learning_rate": 7.3789715468550365e-06, + "loss": 0.0142, + "step": 82340 + }, + { + "epoch": 23.95984870526622, + "grad_norm": 0.10311667621135712, + "learning_rate": 7.375411433661281e-06, + "loss": 0.0128, + "step": 82350 + }, + { + "epoch": 23.962758219377363, + "grad_norm": 0.1287662386894226, + "learning_rate": 7.371851899504712e-06, + "loss": 0.012, + "step": 82360 + }, + { + "epoch": 23.96566773348851, + "grad_norm": 0.09995871037244797, + "learning_rate": 7.368292944655653e-06, + "loss": 0.0133, + "step": 82370 + }, + { + "epoch": 23.96857724759965, + "grad_norm": 0.07809341698884964, + "learning_rate": 7.3647345693843746e-06, + "loss": 0.0122, + "step": 82380 + }, + { + "epoch": 23.971486761710793, + "grad_norm": 0.11243429780006409, + "learning_rate": 7.361176773961118e-06, + "loss": 0.0117, + "step": 82390 + }, + { + "epoch": 23.974396275821938, + "grad_norm": 0.16213670372962952, + "learning_rate": 7.357619558656079e-06, + "loss": 0.012, + "step": 82400 + }, + { + "epoch": 23.97730578993308, + "grad_norm": 0.09824518859386444, + "learning_rate": 7.3540629237394e-06, + "loss": 0.0154, + "step": 82410 + }, + { + "epoch": 23.980215304044226, + "grad_norm": 0.17611005902290344, + "learning_rate": 7.350506869481177e-06, + "loss": 0.0122, + "step": 82420 + }, + { + "epoch": 23.983124818155368, + "grad_norm": 0.13222116231918335, + "learning_rate": 7.3469513961514745e-06, + "loss": 0.0128, + "step": 82430 + }, + { + "epoch": 23.986034332266513, + "grad_norm": 0.07638631016016006, + "learning_rate": 7.343396504020313e-06, + "loss": 0.0112, + "step": 82440 + }, + { + "epoch": 23.988943846377655, + "grad_norm": 0.15621493756771088, + "learning_rate": 7.339842193357659e-06, + "loss": 0.0142, + "step": 82450 + }, + { + "epoch": 23.991853360488797, + "grad_norm": 0.08129988610744476, + "learning_rate": 7.336288464433427e-06, + "loss": 0.0116, + "step": 82460 + }, + { + "epoch": 23.994762874599942, + "grad_norm": 0.09794914722442627, + "learning_rate": 7.332735317517519e-06, + "loss": 0.0122, + "step": 82470 + }, + { + "epoch": 23.997672388711084, + "grad_norm": 0.10776340961456299, + "learning_rate": 7.329182752879765e-06, + "loss": 0.0131, + "step": 82480 + }, + { + "epoch": 24.00058190282223, + "grad_norm": 0.14134608209133148, + "learning_rate": 7.3256307707899525e-06, + "loss": 0.0125, + "step": 82490 + }, + { + "epoch": 24.00349141693337, + "grad_norm": 0.11423341929912567, + "learning_rate": 7.322079371517838e-06, + "loss": 0.0133, + "step": 82500 + }, + { + "epoch": 24.006400931044517, + "grad_norm": 0.09195499122142792, + "learning_rate": 7.31852855533313e-06, + "loss": 0.0123, + "step": 82510 + }, + { + "epoch": 24.00931044515566, + "grad_norm": 0.12623442709445953, + "learning_rate": 7.314978322505485e-06, + "loss": 0.0122, + "step": 82520 + }, + { + "epoch": 24.0122199592668, + "grad_norm": 0.09605465084314346, + "learning_rate": 7.311428673304513e-06, + "loss": 0.0122, + "step": 82530 + }, + { + "epoch": 24.015129473377947, + "grad_norm": 0.1834714114665985, + "learning_rate": 7.307879607999794e-06, + "loss": 0.0136, + "step": 82540 + }, + { + "epoch": 24.01803898748909, + "grad_norm": 0.10073646157979965, + "learning_rate": 7.304331126860859e-06, + "loss": 0.0131, + "step": 82550 + }, + { + "epoch": 24.020948501600234, + "grad_norm": 0.09505777060985565, + "learning_rate": 7.300783230157185e-06, + "loss": 0.0119, + "step": 82560 + }, + { + "epoch": 24.023858015711376, + "grad_norm": 0.11527670174837112, + "learning_rate": 7.297235918158213e-06, + "loss": 0.0128, + "step": 82570 + }, + { + "epoch": 24.02676752982252, + "grad_norm": 0.10394103080034256, + "learning_rate": 7.293689191133345e-06, + "loss": 0.0122, + "step": 82580 + }, + { + "epoch": 24.029677043933663, + "grad_norm": 0.11201265454292297, + "learning_rate": 7.290143049351926e-06, + "loss": 0.0135, + "step": 82590 + }, + { + "epoch": 24.032586558044805, + "grad_norm": 0.1349404901266098, + "learning_rate": 7.286597493083255e-06, + "loss": 0.0148, + "step": 82600 + }, + { + "epoch": 24.03549607215595, + "grad_norm": 0.10768667608499527, + "learning_rate": 7.283052522596601e-06, + "loss": 0.0123, + "step": 82610 + }, + { + "epoch": 24.038405586267093, + "grad_norm": 0.09394870698451996, + "learning_rate": 7.279508138161188e-06, + "loss": 0.0135, + "step": 82620 + }, + { + "epoch": 24.041315100378238, + "grad_norm": 0.1280023753643036, + "learning_rate": 7.275964340046174e-06, + "loss": 0.014, + "step": 82630 + }, + { + "epoch": 24.04422461448938, + "grad_norm": 0.12544430792331696, + "learning_rate": 7.272421128520698e-06, + "loss": 0.0128, + "step": 82640 + }, + { + "epoch": 24.047134128600522, + "grad_norm": 0.11309689283370972, + "learning_rate": 7.268878503853835e-06, + "loss": 0.0125, + "step": 82650 + }, + { + "epoch": 24.050043642711668, + "grad_norm": 0.09666505455970764, + "learning_rate": 7.265336466314631e-06, + "loss": 0.0119, + "step": 82660 + }, + { + "epoch": 24.05295315682281, + "grad_norm": 0.09268681704998016, + "learning_rate": 7.261795016172082e-06, + "loss": 0.0125, + "step": 82670 + }, + { + "epoch": 24.055862670933955, + "grad_norm": 0.10470244288444519, + "learning_rate": 7.258254153695128e-06, + "loss": 0.0122, + "step": 82680 + }, + { + "epoch": 24.058772185045097, + "grad_norm": 0.10435837507247925, + "learning_rate": 7.2547138791526855e-06, + "loss": 0.0173, + "step": 82690 + }, + { + "epoch": 24.061681699156242, + "grad_norm": 0.14002597332000732, + "learning_rate": 7.2511741928136045e-06, + "loss": 0.014, + "step": 82700 + }, + { + "epoch": 24.064591213267384, + "grad_norm": 0.10739494860172272, + "learning_rate": 7.24763509494671e-06, + "loss": 0.013, + "step": 82710 + }, + { + "epoch": 24.067500727378526, + "grad_norm": 0.1249949187040329, + "learning_rate": 7.244096585820764e-06, + "loss": 0.0108, + "step": 82720 + }, + { + "epoch": 24.07041024148967, + "grad_norm": 0.13841913640499115, + "learning_rate": 7.240558665704498e-06, + "loss": 0.0115, + "step": 82730 + }, + { + "epoch": 24.073319755600814, + "grad_norm": 0.11475188285112381, + "learning_rate": 7.237021334866599e-06, + "loss": 0.0119, + "step": 82740 + }, + { + "epoch": 24.07622926971196, + "grad_norm": 0.12822413444519043, + "learning_rate": 7.2334845935757e-06, + "loss": 0.014, + "step": 82750 + }, + { + "epoch": 24.0791387838231, + "grad_norm": 0.11842960119247437, + "learning_rate": 7.229948442100379e-06, + "loss": 0.0129, + "step": 82760 + }, + { + "epoch": 24.082048297934247, + "grad_norm": 0.09881658852100372, + "learning_rate": 7.2264128807092055e-06, + "loss": 0.0115, + "step": 82770 + }, + { + "epoch": 24.08495781204539, + "grad_norm": 0.21406611800193787, + "learning_rate": 7.222877909670674e-06, + "loss": 0.0106, + "step": 82780 + }, + { + "epoch": 24.08786732615653, + "grad_norm": 0.10350848734378815, + "learning_rate": 7.219343529253235e-06, + "loss": 0.0127, + "step": 82790 + }, + { + "epoch": 24.090776840267676, + "grad_norm": 0.11177726835012436, + "learning_rate": 7.215809739725308e-06, + "loss": 0.0132, + "step": 82800 + }, + { + "epoch": 24.093686354378818, + "grad_norm": 0.1693701297044754, + "learning_rate": 7.212276541355266e-06, + "loss": 0.0161, + "step": 82810 + }, + { + "epoch": 24.096595868489963, + "grad_norm": 0.09501361846923828, + "learning_rate": 7.208743934411424e-06, + "loss": 0.0133, + "step": 82820 + }, + { + "epoch": 24.099505382601105, + "grad_norm": 0.11690454185009003, + "learning_rate": 7.205211919162058e-06, + "loss": 0.0115, + "step": 82830 + }, + { + "epoch": 24.10241489671225, + "grad_norm": 0.14387734234333038, + "learning_rate": 7.201680495875405e-06, + "loss": 0.0141, + "step": 82840 + }, + { + "epoch": 24.105324410823393, + "grad_norm": 0.12916497886180878, + "learning_rate": 7.1981496648196606e-06, + "loss": 0.0116, + "step": 82850 + }, + { + "epoch": 24.108233924934535, + "grad_norm": 0.165337935090065, + "learning_rate": 7.194619426262961e-06, + "loss": 0.0116, + "step": 82860 + }, + { + "epoch": 24.11114343904568, + "grad_norm": 0.10101848840713501, + "learning_rate": 7.191089780473393e-06, + "loss": 0.0122, + "step": 82870 + }, + { + "epoch": 24.114052953156822, + "grad_norm": 0.07546888291835785, + "learning_rate": 7.187560727719033e-06, + "loss": 0.0124, + "step": 82880 + }, + { + "epoch": 24.116962467267967, + "grad_norm": 0.1076006069779396, + "learning_rate": 7.184032268267879e-06, + "loss": 0.0147, + "step": 82890 + }, + { + "epoch": 24.11987198137911, + "grad_norm": 0.09824365377426147, + "learning_rate": 7.180504402387888e-06, + "loss": 0.0133, + "step": 82900 + }, + { + "epoch": 24.122781495490255, + "grad_norm": 0.10781807452440262, + "learning_rate": 7.176977130346983e-06, + "loss": 0.0135, + "step": 82910 + }, + { + "epoch": 24.125691009601397, + "grad_norm": 0.11603430658578873, + "learning_rate": 7.173450452413043e-06, + "loss": 0.0119, + "step": 82920 + }, + { + "epoch": 24.12860052371254, + "grad_norm": 0.09247943758964539, + "learning_rate": 7.169924368853884e-06, + "loss": 0.0122, + "step": 82930 + }, + { + "epoch": 24.131510037823684, + "grad_norm": 0.1023123636841774, + "learning_rate": 7.166398879937302e-06, + "loss": 0.012, + "step": 82940 + }, + { + "epoch": 24.134419551934826, + "grad_norm": 0.09903573989868164, + "learning_rate": 7.162873985931023e-06, + "loss": 0.0119, + "step": 82950 + }, + { + "epoch": 24.13732906604597, + "grad_norm": 0.15753623843193054, + "learning_rate": 7.159349687102748e-06, + "loss": 0.0124, + "step": 82960 + }, + { + "epoch": 24.140238580157114, + "grad_norm": 0.09723465144634247, + "learning_rate": 7.155825983720115e-06, + "loss": 0.0117, + "step": 82970 + }, + { + "epoch": 24.143148094268255, + "grad_norm": 0.09602390974760056, + "learning_rate": 7.152302876050733e-06, + "loss": 0.012, + "step": 82980 + }, + { + "epoch": 24.1460576083794, + "grad_norm": 0.12413492798805237, + "learning_rate": 7.148780364362162e-06, + "loss": 0.0141, + "step": 82990 + }, + { + "epoch": 24.148967122490543, + "grad_norm": 0.10741341859102249, + "learning_rate": 7.145258448921903e-06, + "loss": 0.0135, + "step": 83000 + }, + { + "epoch": 24.15187663660169, + "grad_norm": 0.11812126636505127, + "learning_rate": 7.141737129997433e-06, + "loss": 0.0119, + "step": 83010 + }, + { + "epoch": 24.15478615071283, + "grad_norm": 0.15825524926185608, + "learning_rate": 7.1382164078561636e-06, + "loss": 0.0122, + "step": 83020 + }, + { + "epoch": 24.157695664823976, + "grad_norm": 0.074445441365242, + "learning_rate": 7.134696282765473e-06, + "loss": 0.0109, + "step": 83030 + }, + { + "epoch": 24.160605178935118, + "grad_norm": 0.1313764899969101, + "learning_rate": 7.131176754992701e-06, + "loss": 0.0102, + "step": 83040 + }, + { + "epoch": 24.16351469304626, + "grad_norm": 0.12811067700386047, + "learning_rate": 7.1276578248051225e-06, + "loss": 0.0131, + "step": 83050 + }, + { + "epoch": 24.166424207157405, + "grad_norm": 0.09585830569267273, + "learning_rate": 7.124139492469977e-06, + "loss": 0.0115, + "step": 83060 + }, + { + "epoch": 24.169333721268547, + "grad_norm": 0.09484895318746567, + "learning_rate": 7.120621758254458e-06, + "loss": 0.0115, + "step": 83070 + }, + { + "epoch": 24.172243235379693, + "grad_norm": 0.10908060520887375, + "learning_rate": 7.117104622425724e-06, + "loss": 0.012, + "step": 83080 + }, + { + "epoch": 24.175152749490834, + "grad_norm": 0.08403359353542328, + "learning_rate": 7.113588085250865e-06, + "loss": 0.0134, + "step": 83090 + }, + { + "epoch": 24.17806226360198, + "grad_norm": 0.1226910948753357, + "learning_rate": 7.1100721469969445e-06, + "loss": 0.0111, + "step": 83100 + }, + { + "epoch": 24.180971777713122, + "grad_norm": 0.11065986752510071, + "learning_rate": 7.106556807930983e-06, + "loss": 0.0155, + "step": 83110 + }, + { + "epoch": 24.183881291824264, + "grad_norm": 0.08238428086042404, + "learning_rate": 7.103042068319938e-06, + "loss": 0.0112, + "step": 83120 + }, + { + "epoch": 24.18679080593541, + "grad_norm": 0.1462012678384781, + "learning_rate": 7.099527928430727e-06, + "loss": 0.0137, + "step": 83130 + }, + { + "epoch": 24.18970032004655, + "grad_norm": 0.09594299644231796, + "learning_rate": 7.096014388530232e-06, + "loss": 0.0108, + "step": 83140 + }, + { + "epoch": 24.192609834157697, + "grad_norm": 0.10113846510648727, + "learning_rate": 7.092501448885289e-06, + "loss": 0.0116, + "step": 83150 + }, + { + "epoch": 24.19551934826884, + "grad_norm": 0.09511218965053558, + "learning_rate": 7.088989109762675e-06, + "loss": 0.0118, + "step": 83160 + }, + { + "epoch": 24.198428862379984, + "grad_norm": 0.11964470148086548, + "learning_rate": 7.085477371429119e-06, + "loss": 0.0133, + "step": 83170 + }, + { + "epoch": 24.201338376491126, + "grad_norm": 0.1072021871805191, + "learning_rate": 7.081966234151336e-06, + "loss": 0.0133, + "step": 83180 + }, + { + "epoch": 24.204247890602268, + "grad_norm": 0.07359758764505386, + "learning_rate": 7.078455698195963e-06, + "loss": 0.0117, + "step": 83190 + }, + { + "epoch": 24.207157404713413, + "grad_norm": 0.08812130987644196, + "learning_rate": 7.0749457638295975e-06, + "loss": 0.0127, + "step": 83200 + }, + { + "epoch": 24.210066918824555, + "grad_norm": 0.09301616996526718, + "learning_rate": 7.071436431318798e-06, + "loss": 0.011, + "step": 83210 + }, + { + "epoch": 24.2129764329357, + "grad_norm": 0.12970015406608582, + "learning_rate": 7.067927700930085e-06, + "loss": 0.0115, + "step": 83220 + }, + { + "epoch": 24.215885947046843, + "grad_norm": 0.17245851457118988, + "learning_rate": 7.064419572929911e-06, + "loss": 0.0112, + "step": 83230 + }, + { + "epoch": 24.21879546115799, + "grad_norm": 0.07814918458461761, + "learning_rate": 7.060912047584704e-06, + "loss": 0.0125, + "step": 83240 + }, + { + "epoch": 24.22170497526913, + "grad_norm": 0.1193484514951706, + "learning_rate": 7.057405125160829e-06, + "loss": 0.0141, + "step": 83250 + }, + { + "epoch": 24.224614489380272, + "grad_norm": 0.08192596584558487, + "learning_rate": 7.053898805924625e-06, + "loss": 0.0107, + "step": 83260 + }, + { + "epoch": 24.227524003491418, + "grad_norm": 0.14677855372428894, + "learning_rate": 7.0503930901423596e-06, + "loss": 0.0116, + "step": 83270 + }, + { + "epoch": 24.23043351760256, + "grad_norm": 0.1094379872083664, + "learning_rate": 7.046887978080276e-06, + "loss": 0.0116, + "step": 83280 + }, + { + "epoch": 24.233343031713705, + "grad_norm": 0.17159543931484222, + "learning_rate": 7.04338347000457e-06, + "loss": 0.0132, + "step": 83290 + }, + { + "epoch": 24.236252545824847, + "grad_norm": 0.10508960485458374, + "learning_rate": 7.039879566181377e-06, + "loss": 0.0134, + "step": 83300 + }, + { + "epoch": 24.23916205993599, + "grad_norm": 0.09240979701280594, + "learning_rate": 7.036376266876802e-06, + "loss": 0.0141, + "step": 83310 + }, + { + "epoch": 24.242071574047134, + "grad_norm": 0.13323789834976196, + "learning_rate": 7.03287357235689e-06, + "loss": 0.0133, + "step": 83320 + }, + { + "epoch": 24.244981088158276, + "grad_norm": 0.09417569637298584, + "learning_rate": 7.029371482887656e-06, + "loss": 0.0122, + "step": 83330 + }, + { + "epoch": 24.247890602269422, + "grad_norm": 0.11300661414861679, + "learning_rate": 7.025869998735052e-06, + "loss": 0.0118, + "step": 83340 + }, + { + "epoch": 24.250800116380564, + "grad_norm": 0.09162361919879913, + "learning_rate": 7.022369120165003e-06, + "loss": 0.0128, + "step": 83350 + }, + { + "epoch": 24.25370963049171, + "grad_norm": 0.09871980547904968, + "learning_rate": 7.018868847443365e-06, + "loss": 0.0127, + "step": 83360 + }, + { + "epoch": 24.25661914460285, + "grad_norm": 0.15046939253807068, + "learning_rate": 7.0153691808359675e-06, + "loss": 0.0109, + "step": 83370 + }, + { + "epoch": 24.259528658713993, + "grad_norm": 0.10397671908140182, + "learning_rate": 7.01187012060859e-06, + "loss": 0.0132, + "step": 83380 + }, + { + "epoch": 24.26243817282514, + "grad_norm": 0.10557131469249725, + "learning_rate": 7.008371667026958e-06, + "loss": 0.012, + "step": 83390 + }, + { + "epoch": 24.26534768693628, + "grad_norm": 0.10859280824661255, + "learning_rate": 7.004873820356754e-06, + "loss": 0.013, + "step": 83400 + }, + { + "epoch": 24.268257201047426, + "grad_norm": 0.10231134295463562, + "learning_rate": 7.0013765808636295e-06, + "loss": 0.0123, + "step": 83410 + }, + { + "epoch": 24.271166715158568, + "grad_norm": 0.08701319992542267, + "learning_rate": 6.9978799488131635e-06, + "loss": 0.0115, + "step": 83420 + }, + { + "epoch": 24.274076229269713, + "grad_norm": 0.10420706123113632, + "learning_rate": 6.994383924470904e-06, + "loss": 0.0128, + "step": 83430 + }, + { + "epoch": 24.276985743380855, + "grad_norm": 0.08345292508602142, + "learning_rate": 6.990888508102351e-06, + "loss": 0.0106, + "step": 83440 + }, + { + "epoch": 24.279895257491997, + "grad_norm": 0.1129554957151413, + "learning_rate": 6.987393699972966e-06, + "loss": 0.0125, + "step": 83450 + }, + { + "epoch": 24.282804771603143, + "grad_norm": 0.13869084417819977, + "learning_rate": 6.98389950034815e-06, + "loss": 0.0109, + "step": 83460 + }, + { + "epoch": 24.285714285714285, + "grad_norm": 0.10070401430130005, + "learning_rate": 6.980405909493255e-06, + "loss": 0.0119, + "step": 83470 + }, + { + "epoch": 24.28862379982543, + "grad_norm": 0.08557243645191193, + "learning_rate": 6.976912927673619e-06, + "loss": 0.0111, + "step": 83480 + }, + { + "epoch": 24.291533313936572, + "grad_norm": 0.15485282242298126, + "learning_rate": 6.973420555154496e-06, + "loss": 0.0104, + "step": 83490 + }, + { + "epoch": 24.294442828047718, + "grad_norm": 0.21515241265296936, + "learning_rate": 6.969928792201108e-06, + "loss": 0.0129, + "step": 83500 + }, + { + "epoch": 24.29735234215886, + "grad_norm": 0.11799183487892151, + "learning_rate": 6.966437639078632e-06, + "loss": 0.0141, + "step": 83510 + }, + { + "epoch": 24.30026185627, + "grad_norm": 0.07658097892999649, + "learning_rate": 6.9629470960522065e-06, + "loss": 0.0111, + "step": 83520 + }, + { + "epoch": 24.303171370381147, + "grad_norm": 0.08752334117889404, + "learning_rate": 6.95945716338691e-06, + "loss": 0.012, + "step": 83530 + }, + { + "epoch": 24.30608088449229, + "grad_norm": 0.10720223933458328, + "learning_rate": 6.955967841347772e-06, + "loss": 0.0126, + "step": 83540 + }, + { + "epoch": 24.308990398603434, + "grad_norm": 0.12938199937343597, + "learning_rate": 6.9524791301997924e-06, + "loss": 0.011, + "step": 83550 + }, + { + "epoch": 24.311899912714576, + "grad_norm": 0.11460397392511368, + "learning_rate": 6.948991030207918e-06, + "loss": 0.013, + "step": 83560 + }, + { + "epoch": 24.31480942682572, + "grad_norm": 0.09389664232730865, + "learning_rate": 6.945503541637038e-06, + "loss": 0.0116, + "step": 83570 + }, + { + "epoch": 24.317718940936864, + "grad_norm": 0.1028117761015892, + "learning_rate": 6.942016664752009e-06, + "loss": 0.0101, + "step": 83580 + }, + { + "epoch": 24.320628455048006, + "grad_norm": 0.1144455298781395, + "learning_rate": 6.938530399817642e-06, + "loss": 0.0127, + "step": 83590 + }, + { + "epoch": 24.32353796915915, + "grad_norm": 0.09642479568719864, + "learning_rate": 6.935044747098684e-06, + "loss": 0.013, + "step": 83600 + }, + { + "epoch": 24.326447483270293, + "grad_norm": 0.18028739094734192, + "learning_rate": 6.931559706859857e-06, + "loss": 0.0106, + "step": 83610 + }, + { + "epoch": 24.32935699738144, + "grad_norm": 0.10355337709188461, + "learning_rate": 6.928075279365821e-06, + "loss": 0.012, + "step": 83620 + }, + { + "epoch": 24.33226651149258, + "grad_norm": 0.10301706939935684, + "learning_rate": 6.924591464881201e-06, + "loss": 0.0139, + "step": 83630 + }, + { + "epoch": 24.335176025603722, + "grad_norm": 0.0841703861951828, + "learning_rate": 6.921108263670561e-06, + "loss": 0.011, + "step": 83640 + }, + { + "epoch": 24.338085539714868, + "grad_norm": 0.11399044841527939, + "learning_rate": 6.917625675998439e-06, + "loss": 0.0112, + "step": 83650 + }, + { + "epoch": 24.34099505382601, + "grad_norm": 0.09500688314437866, + "learning_rate": 6.914143702129301e-06, + "loss": 0.0119, + "step": 83660 + }, + { + "epoch": 24.343904567937155, + "grad_norm": 0.10691593587398529, + "learning_rate": 6.910662342327588e-06, + "loss": 0.0137, + "step": 83670 + }, + { + "epoch": 24.346814082048297, + "grad_norm": 0.16883257031440735, + "learning_rate": 6.907181596857688e-06, + "loss": 0.0129, + "step": 83680 + }, + { + "epoch": 24.349723596159443, + "grad_norm": 0.14022555947303772, + "learning_rate": 6.903701465983935e-06, + "loss": 0.0126, + "step": 83690 + }, + { + "epoch": 24.352633110270585, + "grad_norm": 0.11728333681821823, + "learning_rate": 6.900221949970629e-06, + "loss": 0.0114, + "step": 83700 + }, + { + "epoch": 24.355542624381727, + "grad_norm": 0.11998418718576431, + "learning_rate": 6.896743049082007e-06, + "loss": 0.0127, + "step": 83710 + }, + { + "epoch": 24.358452138492872, + "grad_norm": 0.11773742735385895, + "learning_rate": 6.893264763582279e-06, + "loss": 0.0136, + "step": 83720 + }, + { + "epoch": 24.361361652604014, + "grad_norm": 0.16111917793750763, + "learning_rate": 6.889787093735587e-06, + "loss": 0.0119, + "step": 83730 + }, + { + "epoch": 24.36427116671516, + "grad_norm": 0.08807458728551865, + "learning_rate": 6.886310039806043e-06, + "loss": 0.0104, + "step": 83740 + }, + { + "epoch": 24.3671806808263, + "grad_norm": 0.0880085676908493, + "learning_rate": 6.88283360205771e-06, + "loss": 0.0114, + "step": 83750 + }, + { + "epoch": 24.370090194937447, + "grad_norm": 0.09356711804866791, + "learning_rate": 6.8793577807545966e-06, + "loss": 0.0116, + "step": 83760 + }, + { + "epoch": 24.37299970904859, + "grad_norm": 0.1082274317741394, + "learning_rate": 6.875882576160657e-06, + "loss": 0.0124, + "step": 83770 + }, + { + "epoch": 24.37590922315973, + "grad_norm": 0.0960974246263504, + "learning_rate": 6.872407988539833e-06, + "loss": 0.0131, + "step": 83780 + }, + { + "epoch": 24.378818737270876, + "grad_norm": 0.11708246171474457, + "learning_rate": 6.8689340181559844e-06, + "loss": 0.013, + "step": 83790 + }, + { + "epoch": 24.381728251382018, + "grad_norm": 0.11301849037408829, + "learning_rate": 6.8654606652729315e-06, + "loss": 0.0116, + "step": 83800 + }, + { + "epoch": 24.384637765493164, + "grad_norm": 0.09922720491886139, + "learning_rate": 6.861987930154457e-06, + "loss": 0.0106, + "step": 83810 + }, + { + "epoch": 24.387547279604306, + "grad_norm": 0.12191013246774673, + "learning_rate": 6.8585158130643e-06, + "loss": 0.011, + "step": 83820 + }, + { + "epoch": 24.39045679371545, + "grad_norm": 0.12471474707126617, + "learning_rate": 6.855044314266138e-06, + "loss": 0.0123, + "step": 83830 + }, + { + "epoch": 24.393366307826593, + "grad_norm": 0.13783583045005798, + "learning_rate": 6.851573434023602e-06, + "loss": 0.0105, + "step": 83840 + }, + { + "epoch": 24.396275821937735, + "grad_norm": 0.12884919345378876, + "learning_rate": 6.848103172600287e-06, + "loss": 0.0127, + "step": 83850 + }, + { + "epoch": 24.39918533604888, + "grad_norm": 0.08345529437065125, + "learning_rate": 6.844633530259747e-06, + "loss": 0.0125, + "step": 83860 + }, + { + "epoch": 24.402094850160022, + "grad_norm": 0.09292532503604889, + "learning_rate": 6.841164507265461e-06, + "loss": 0.0116, + "step": 83870 + }, + { + "epoch": 24.405004364271168, + "grad_norm": 0.0853348895907402, + "learning_rate": 6.8376961038808886e-06, + "loss": 0.0132, + "step": 83880 + }, + { + "epoch": 24.40791387838231, + "grad_norm": 0.09345197677612305, + "learning_rate": 6.834228320369436e-06, + "loss": 0.0132, + "step": 83890 + }, + { + "epoch": 24.410823392493455, + "grad_norm": 0.10964822769165039, + "learning_rate": 6.830761156994454e-06, + "loss": 0.0123, + "step": 83900 + }, + { + "epoch": 24.413732906604597, + "grad_norm": 0.07953955233097076, + "learning_rate": 6.827294614019243e-06, + "loss": 0.0112, + "step": 83910 + }, + { + "epoch": 24.41664242071574, + "grad_norm": 0.09500192105770111, + "learning_rate": 6.82382869170707e-06, + "loss": 0.0109, + "step": 83920 + }, + { + "epoch": 24.419551934826885, + "grad_norm": 0.13088753819465637, + "learning_rate": 6.820363390321158e-06, + "loss": 0.0117, + "step": 83930 + }, + { + "epoch": 24.422461448938027, + "grad_norm": 0.10180121660232544, + "learning_rate": 6.816898710124658e-06, + "loss": 0.0124, + "step": 83940 + }, + { + "epoch": 24.425370963049172, + "grad_norm": 0.1419447809457779, + "learning_rate": 6.8134346513807035e-06, + "loss": 0.0121, + "step": 83950 + }, + { + "epoch": 24.428280477160314, + "grad_norm": 0.13770371675491333, + "learning_rate": 6.8099712143523555e-06, + "loss": 0.0135, + "step": 83960 + }, + { + "epoch": 24.43118999127146, + "grad_norm": 0.07930774986743927, + "learning_rate": 6.806508399302644e-06, + "loss": 0.0113, + "step": 83970 + }, + { + "epoch": 24.4340995053826, + "grad_norm": 0.10922575742006302, + "learning_rate": 6.803046206494553e-06, + "loss": 0.0142, + "step": 83980 + }, + { + "epoch": 24.437009019493743, + "grad_norm": 0.10654281824827194, + "learning_rate": 6.7995846361910015e-06, + "loss": 0.0134, + "step": 83990 + }, + { + "epoch": 24.43991853360489, + "grad_norm": 0.12408339232206345, + "learning_rate": 6.796123688654883e-06, + "loss": 0.0115, + "step": 84000 + }, + { + "epoch": 24.44282804771603, + "grad_norm": 0.08873133361339569, + "learning_rate": 6.792663364149025e-06, + "loss": 0.0109, + "step": 84010 + }, + { + "epoch": 24.445737561827176, + "grad_norm": 0.092460498213768, + "learning_rate": 6.789203662936227e-06, + "loss": 0.0125, + "step": 84020 + }, + { + "epoch": 24.448647075938318, + "grad_norm": 0.13622546195983887, + "learning_rate": 6.785744585279219e-06, + "loss": 0.0137, + "step": 84030 + }, + { + "epoch": 24.45155659004946, + "grad_norm": 0.12668123841285706, + "learning_rate": 6.7822861314406986e-06, + "loss": 0.0118, + "step": 84040 + }, + { + "epoch": 24.454466104160606, + "grad_norm": 0.08967998623847961, + "learning_rate": 6.778828301683322e-06, + "loss": 0.0129, + "step": 84050 + }, + { + "epoch": 24.457375618271747, + "grad_norm": 0.14119884371757507, + "learning_rate": 6.775371096269681e-06, + "loss": 0.0162, + "step": 84060 + }, + { + "epoch": 24.460285132382893, + "grad_norm": 0.10737556964159012, + "learning_rate": 6.771914515462319e-06, + "loss": 0.0106, + "step": 84070 + }, + { + "epoch": 24.463194646494035, + "grad_norm": 0.09337040036916733, + "learning_rate": 6.76845855952375e-06, + "loss": 0.0107, + "step": 84080 + }, + { + "epoch": 24.46610416060518, + "grad_norm": 0.10821323841810226, + "learning_rate": 6.765003228716437e-06, + "loss": 0.012, + "step": 84090 + }, + { + "epoch": 24.469013674716322, + "grad_norm": 0.1224534884095192, + "learning_rate": 6.761548523302776e-06, + "loss": 0.012, + "step": 84100 + }, + { + "epoch": 24.471923188827464, + "grad_norm": 0.12908561527729034, + "learning_rate": 6.758094443545136e-06, + "loss": 0.0109, + "step": 84110 + }, + { + "epoch": 24.47483270293861, + "grad_norm": 0.1298784762620926, + "learning_rate": 6.754640989705836e-06, + "loss": 0.0135, + "step": 84120 + }, + { + "epoch": 24.47774221704975, + "grad_norm": 0.10730090737342834, + "learning_rate": 6.75118816204714e-06, + "loss": 0.0124, + "step": 84130 + }, + { + "epoch": 24.480651731160897, + "grad_norm": 0.06932874768972397, + "learning_rate": 6.747735960831258e-06, + "loss": 0.011, + "step": 84140 + }, + { + "epoch": 24.48356124527204, + "grad_norm": 0.10484591126441956, + "learning_rate": 6.74428438632037e-06, + "loss": 0.0127, + "step": 84150 + }, + { + "epoch": 24.486470759383185, + "grad_norm": 0.0943303033709526, + "learning_rate": 6.740833438776605e-06, + "loss": 0.0119, + "step": 84160 + }, + { + "epoch": 24.489380273494326, + "grad_norm": 0.1952827125787735, + "learning_rate": 6.737383118462036e-06, + "loss": 0.0112, + "step": 84170 + }, + { + "epoch": 24.49228978760547, + "grad_norm": 0.1292068064212799, + "learning_rate": 6.733933425638679e-06, + "loss": 0.013, + "step": 84180 + }, + { + "epoch": 24.495199301716614, + "grad_norm": 0.09891398996114731, + "learning_rate": 6.730484360568537e-06, + "loss": 0.0152, + "step": 84190 + }, + { + "epoch": 24.498108815827756, + "grad_norm": 0.10014667361974716, + "learning_rate": 6.727035923513536e-06, + "loss": 0.0131, + "step": 84200 + }, + { + "epoch": 24.5010183299389, + "grad_norm": 0.09826000779867172, + "learning_rate": 6.723588114735552e-06, + "loss": 0.0118, + "step": 84210 + }, + { + "epoch": 24.503927844050043, + "grad_norm": 0.12038980424404144, + "learning_rate": 6.720140934496432e-06, + "loss": 0.0125, + "step": 84220 + }, + { + "epoch": 24.50683735816119, + "grad_norm": 0.09138856828212738, + "learning_rate": 6.716694383057971e-06, + "loss": 0.0114, + "step": 84230 + }, + { + "epoch": 24.50974687227233, + "grad_norm": 0.18325775861740112, + "learning_rate": 6.713248460681903e-06, + "loss": 0.0132, + "step": 84240 + }, + { + "epoch": 24.512656386383473, + "grad_norm": 0.16167372465133667, + "learning_rate": 6.709803167629929e-06, + "loss": 0.0122, + "step": 84250 + }, + { + "epoch": 24.515565900494618, + "grad_norm": 0.12649112939834595, + "learning_rate": 6.70635850416369e-06, + "loss": 0.0142, + "step": 84260 + }, + { + "epoch": 24.51847541460576, + "grad_norm": 0.11901970952749252, + "learning_rate": 6.702914470544795e-06, + "loss": 0.0111, + "step": 84270 + }, + { + "epoch": 24.521384928716905, + "grad_norm": 0.11200658231973648, + "learning_rate": 6.699471067034784e-06, + "loss": 0.0106, + "step": 84280 + }, + { + "epoch": 24.524294442828047, + "grad_norm": 0.11619196087121964, + "learning_rate": 6.696028293895167e-06, + "loss": 0.0102, + "step": 84290 + }, + { + "epoch": 24.527203956939193, + "grad_norm": 0.08826770633459091, + "learning_rate": 6.6925861513874044e-06, + "loss": 0.0118, + "step": 84300 + }, + { + "epoch": 24.530113471050335, + "grad_norm": 0.12257298082113266, + "learning_rate": 6.689144639772895e-06, + "loss": 0.0128, + "step": 84310 + }, + { + "epoch": 24.533022985161477, + "grad_norm": 0.12080123275518417, + "learning_rate": 6.685703759313009e-06, + "loss": 0.0132, + "step": 84320 + }, + { + "epoch": 24.535932499272622, + "grad_norm": 0.0904434472322464, + "learning_rate": 6.682263510269045e-06, + "loss": 0.0109, + "step": 84330 + }, + { + "epoch": 24.538842013383764, + "grad_norm": 0.10573882609605789, + "learning_rate": 6.6788238929022775e-06, + "loss": 0.0119, + "step": 84340 + }, + { + "epoch": 24.54175152749491, + "grad_norm": 0.11669043451547623, + "learning_rate": 6.675384907473927e-06, + "loss": 0.0102, + "step": 84350 + }, + { + "epoch": 24.54466104160605, + "grad_norm": 0.1005919948220253, + "learning_rate": 6.671946554245153e-06, + "loss": 0.0125, + "step": 84360 + }, + { + "epoch": 24.547570555717193, + "grad_norm": 0.11740623414516449, + "learning_rate": 6.668508833477072e-06, + "loss": 0.013, + "step": 84370 + }, + { + "epoch": 24.55048006982834, + "grad_norm": 0.12094961106777191, + "learning_rate": 6.665071745430762e-06, + "loss": 0.0107, + "step": 84380 + }, + { + "epoch": 24.55338958393948, + "grad_norm": 0.06900829076766968, + "learning_rate": 6.661635290367254e-06, + "loss": 0.0106, + "step": 84390 + }, + { + "epoch": 24.556299098050626, + "grad_norm": 0.11806595325469971, + "learning_rate": 6.6581994685475105e-06, + "loss": 0.0107, + "step": 84400 + }, + { + "epoch": 24.55920861216177, + "grad_norm": 0.09764142334461212, + "learning_rate": 6.6547642802324684e-06, + "loss": 0.0103, + "step": 84410 + }, + { + "epoch": 24.562118126272914, + "grad_norm": 0.0978640541434288, + "learning_rate": 6.651329725683009e-06, + "loss": 0.0104, + "step": 84420 + }, + { + "epoch": 24.565027640384056, + "grad_norm": 0.08588864654302597, + "learning_rate": 6.647895805159962e-06, + "loss": 0.0103, + "step": 84430 + }, + { + "epoch": 24.567937154495198, + "grad_norm": 0.2674052119255066, + "learning_rate": 6.644462518924096e-06, + "loss": 0.0138, + "step": 84440 + }, + { + "epoch": 24.570846668606343, + "grad_norm": 0.10221811383962631, + "learning_rate": 6.641029867236173e-06, + "loss": 0.0132, + "step": 84450 + }, + { + "epoch": 24.573756182717485, + "grad_norm": 0.1357974261045456, + "learning_rate": 6.637597850356867e-06, + "loss": 0.0122, + "step": 84460 + }, + { + "epoch": 24.57666569682863, + "grad_norm": 0.098868228495121, + "learning_rate": 6.634166468546816e-06, + "loss": 0.0108, + "step": 84470 + }, + { + "epoch": 24.579575210939772, + "grad_norm": 0.05891802906990051, + "learning_rate": 6.630735722066602e-06, + "loss": 0.013, + "step": 84480 + }, + { + "epoch": 24.582484725050918, + "grad_norm": 0.21239420771598816, + "learning_rate": 6.627305611176788e-06, + "loss": 0.0114, + "step": 84490 + }, + { + "epoch": 24.58539423916206, + "grad_norm": 0.09122376143932343, + "learning_rate": 6.62387613613786e-06, + "loss": 0.0121, + "step": 84500 + }, + { + "epoch": 24.588303753273202, + "grad_norm": 0.09854885190725327, + "learning_rate": 6.620447297210253e-06, + "loss": 0.0121, + "step": 84510 + }, + { + "epoch": 24.591213267384347, + "grad_norm": 0.17235167324543, + "learning_rate": 6.6170190946543765e-06, + "loss": 0.0118, + "step": 84520 + }, + { + "epoch": 24.59412278149549, + "grad_norm": 0.16576963663101196, + "learning_rate": 6.613591528730581e-06, + "loss": 0.0132, + "step": 84530 + }, + { + "epoch": 24.597032295606635, + "grad_norm": 0.08923134952783585, + "learning_rate": 6.610164599699164e-06, + "loss": 0.0128, + "step": 84540 + }, + { + "epoch": 24.599941809717777, + "grad_norm": 0.08264674246311188, + "learning_rate": 6.606738307820366e-06, + "loss": 0.0104, + "step": 84550 + }, + { + "epoch": 24.602851323828922, + "grad_norm": 0.09061755239963531, + "learning_rate": 6.603312653354414e-06, + "loss": 0.0108, + "step": 84560 + }, + { + "epoch": 24.605760837940064, + "grad_norm": 0.09853769838809967, + "learning_rate": 6.599887636561452e-06, + "loss": 0.011, + "step": 84570 + }, + { + "epoch": 24.608670352051206, + "grad_norm": 0.09948179870843887, + "learning_rate": 6.5964632577015824e-06, + "loss": 0.0125, + "step": 84580 + }, + { + "epoch": 24.61157986616235, + "grad_norm": 0.16983796656131744, + "learning_rate": 6.593039517034871e-06, + "loss": 0.0125, + "step": 84590 + }, + { + "epoch": 24.614489380273493, + "grad_norm": 0.19061626493930817, + "learning_rate": 6.5896164148213314e-06, + "loss": 0.0112, + "step": 84600 + }, + { + "epoch": 24.61739889438464, + "grad_norm": 0.10366039723157883, + "learning_rate": 6.586193951320918e-06, + "loss": 0.0117, + "step": 84610 + }, + { + "epoch": 24.62030840849578, + "grad_norm": 0.15226268768310547, + "learning_rate": 6.582772126793552e-06, + "loss": 0.012, + "step": 84620 + }, + { + "epoch": 24.623217922606926, + "grad_norm": 0.08066537976264954, + "learning_rate": 6.579350941499088e-06, + "loss": 0.0125, + "step": 84630 + }, + { + "epoch": 24.62612743671807, + "grad_norm": 0.10176479816436768, + "learning_rate": 6.575930395697356e-06, + "loss": 0.0125, + "step": 84640 + }, + { + "epoch": 24.62903695082921, + "grad_norm": 0.09698158502578735, + "learning_rate": 6.572510489648108e-06, + "loss": 0.0141, + "step": 84650 + }, + { + "epoch": 24.631946464940356, + "grad_norm": 0.13663044571876526, + "learning_rate": 6.569091223611073e-06, + "loss": 0.0137, + "step": 84660 + }, + { + "epoch": 24.634855979051498, + "grad_norm": 0.10159241408109665, + "learning_rate": 6.565672597845927e-06, + "loss": 0.0126, + "step": 84670 + }, + { + "epoch": 24.637765493162643, + "grad_norm": 0.07903256267309189, + "learning_rate": 6.562254612612279e-06, + "loss": 0.0117, + "step": 84680 + }, + { + "epoch": 24.640675007273785, + "grad_norm": 0.09893683344125748, + "learning_rate": 6.558837268169715e-06, + "loss": 0.0125, + "step": 84690 + }, + { + "epoch": 24.64358452138493, + "grad_norm": 0.11926772445440292, + "learning_rate": 6.555420564777748e-06, + "loss": 0.0138, + "step": 84700 + }, + { + "epoch": 24.646494035496072, + "grad_norm": 0.10656420886516571, + "learning_rate": 6.552004502695859e-06, + "loss": 0.0119, + "step": 84710 + }, + { + "epoch": 24.649403549607214, + "grad_norm": 0.13725194334983826, + "learning_rate": 6.548589082183481e-06, + "loss": 0.0124, + "step": 84720 + }, + { + "epoch": 24.65231306371836, + "grad_norm": 0.09096939116716385, + "learning_rate": 6.545174303499989e-06, + "loss": 0.014, + "step": 84730 + }, + { + "epoch": 24.6552225778295, + "grad_norm": 0.07673488557338715, + "learning_rate": 6.541760166904705e-06, + "loss": 0.0114, + "step": 84740 + }, + { + "epoch": 24.658132091940647, + "grad_norm": 0.08832289278507233, + "learning_rate": 6.538346672656919e-06, + "loss": 0.0116, + "step": 84750 + }, + { + "epoch": 24.66104160605179, + "grad_norm": 0.14008720219135284, + "learning_rate": 6.5349338210158654e-06, + "loss": 0.0122, + "step": 84760 + }, + { + "epoch": 24.66395112016293, + "grad_norm": 0.10672137886285782, + "learning_rate": 6.5315216122407176e-06, + "loss": 0.0112, + "step": 84770 + }, + { + "epoch": 24.666860634274077, + "grad_norm": 0.1258135437965393, + "learning_rate": 6.528110046590617e-06, + "loss": 0.0136, + "step": 84780 + }, + { + "epoch": 24.66977014838522, + "grad_norm": 0.12626062333583832, + "learning_rate": 6.5246991243246535e-06, + "loss": 0.0127, + "step": 84790 + }, + { + "epoch": 24.672679662496364, + "grad_norm": 0.0899471864104271, + "learning_rate": 6.52128884570186e-06, + "loss": 0.0107, + "step": 84800 + }, + { + "epoch": 24.675589176607506, + "grad_norm": 0.08660640567541122, + "learning_rate": 6.517879210981218e-06, + "loss": 0.0129, + "step": 84810 + }, + { + "epoch": 24.67849869071865, + "grad_norm": 0.14390166103839874, + "learning_rate": 6.514470220421674e-06, + "loss": 0.0124, + "step": 84820 + }, + { + "epoch": 24.681408204829793, + "grad_norm": 0.0861000046133995, + "learning_rate": 6.511061874282122e-06, + "loss": 0.0146, + "step": 84830 + }, + { + "epoch": 24.684317718940935, + "grad_norm": 0.08985184133052826, + "learning_rate": 6.507654172821397e-06, + "loss": 0.0119, + "step": 84840 + }, + { + "epoch": 24.68722723305208, + "grad_norm": 0.10922583192586899, + "learning_rate": 6.504247116298284e-06, + "loss": 0.0127, + "step": 84850 + }, + { + "epoch": 24.690136747163223, + "grad_norm": 0.08032093942165375, + "learning_rate": 6.500840704971546e-06, + "loss": 0.0109, + "step": 84860 + }, + { + "epoch": 24.693046261274368, + "grad_norm": 0.1039203554391861, + "learning_rate": 6.4974349390998665e-06, + "loss": 0.0119, + "step": 84870 + }, + { + "epoch": 24.69595577538551, + "grad_norm": 0.1324070692062378, + "learning_rate": 6.494029818941885e-06, + "loss": 0.0115, + "step": 84880 + }, + { + "epoch": 24.698865289496656, + "grad_norm": 0.08818351477384567, + "learning_rate": 6.490625344756206e-06, + "loss": 0.0111, + "step": 84890 + }, + { + "epoch": 24.701774803607798, + "grad_norm": 0.1277056336402893, + "learning_rate": 6.4872215168013804e-06, + "loss": 0.0119, + "step": 84900 + }, + { + "epoch": 24.70468431771894, + "grad_norm": 0.12060955911874771, + "learning_rate": 6.4838183353359e-06, + "loss": 0.0131, + "step": 84910 + }, + { + "epoch": 24.707593831830085, + "grad_norm": 0.09583266824483871, + "learning_rate": 6.480415800618211e-06, + "loss": 0.011, + "step": 84920 + }, + { + "epoch": 24.710503345941227, + "grad_norm": 0.11352371424436569, + "learning_rate": 6.477013912906716e-06, + "loss": 0.0113, + "step": 84930 + }, + { + "epoch": 24.713412860052372, + "grad_norm": 0.1432604044675827, + "learning_rate": 6.473612672459774e-06, + "loss": 0.0118, + "step": 84940 + }, + { + "epoch": 24.716322374163514, + "grad_norm": 0.1313677281141281, + "learning_rate": 6.470212079535674e-06, + "loss": 0.0127, + "step": 84950 + }, + { + "epoch": 24.71923188827466, + "grad_norm": 0.09042922407388687, + "learning_rate": 6.466812134392675e-06, + "loss": 0.011, + "step": 84960 + }, + { + "epoch": 24.7221414023858, + "grad_norm": 0.12830260396003723, + "learning_rate": 6.463412837288985e-06, + "loss": 0.0118, + "step": 84970 + }, + { + "epoch": 24.725050916496944, + "grad_norm": 0.1021689772605896, + "learning_rate": 6.4600141884827505e-06, + "loss": 0.0111, + "step": 84980 + }, + { + "epoch": 24.72796043060809, + "grad_norm": 0.1302712857723236, + "learning_rate": 6.456616188232082e-06, + "loss": 0.0117, + "step": 84990 + }, + { + "epoch": 24.73086994471923, + "grad_norm": 0.078711599111557, + "learning_rate": 6.453218836795029e-06, + "loss": 0.0123, + "step": 85000 + }, + { + "epoch": 24.733779458830377, + "grad_norm": 0.07898242771625519, + "learning_rate": 6.449822134429607e-06, + "loss": 0.012, + "step": 85010 + }, + { + "epoch": 24.73668897294152, + "grad_norm": 0.14702893793582916, + "learning_rate": 6.446426081393762e-06, + "loss": 0.0124, + "step": 85020 + }, + { + "epoch": 24.73959848705266, + "grad_norm": 0.16647924482822418, + "learning_rate": 6.443030677945413e-06, + "loss": 0.0158, + "step": 85030 + }, + { + "epoch": 24.742508001163806, + "grad_norm": 0.11932359635829926, + "learning_rate": 6.439635924342405e-06, + "loss": 0.0114, + "step": 85040 + }, + { + "epoch": 24.745417515274948, + "grad_norm": 0.10614825040102005, + "learning_rate": 6.436241820842558e-06, + "loss": 0.0128, + "step": 85050 + }, + { + "epoch": 24.748327029386093, + "grad_norm": 0.1117333173751831, + "learning_rate": 6.432848367703634e-06, + "loss": 0.012, + "step": 85060 + }, + { + "epoch": 24.751236543497235, + "grad_norm": 0.11020760983228683, + "learning_rate": 6.429455565183332e-06, + "loss": 0.013, + "step": 85070 + }, + { + "epoch": 24.75414605760838, + "grad_norm": 0.2211574912071228, + "learning_rate": 6.426063413539318e-06, + "loss": 0.0114, + "step": 85080 + }, + { + "epoch": 24.757055571719523, + "grad_norm": 0.1182866096496582, + "learning_rate": 6.422671913029213e-06, + "loss": 0.011, + "step": 85090 + }, + { + "epoch": 24.759965085830665, + "grad_norm": 0.1298506259918213, + "learning_rate": 6.419281063910568e-06, + "loss": 0.0112, + "step": 85100 + }, + { + "epoch": 24.76287459994181, + "grad_norm": 0.10450176894664764, + "learning_rate": 6.415890866440893e-06, + "loss": 0.0124, + "step": 85110 + }, + { + "epoch": 24.765784114052952, + "grad_norm": 0.13375215232372284, + "learning_rate": 6.412501320877659e-06, + "loss": 0.0106, + "step": 85120 + }, + { + "epoch": 24.768693628164097, + "grad_norm": 0.12171649187803268, + "learning_rate": 6.409112427478281e-06, + "loss": 0.013, + "step": 85130 + }, + { + "epoch": 24.77160314227524, + "grad_norm": 0.12570516765117645, + "learning_rate": 6.4057241865001185e-06, + "loss": 0.0124, + "step": 85140 + }, + { + "epoch": 24.774512656386385, + "grad_norm": 0.08737924695014954, + "learning_rate": 6.402336598200476e-06, + "loss": 0.0113, + "step": 85150 + }, + { + "epoch": 24.777422170497527, + "grad_norm": 0.12750956416130066, + "learning_rate": 6.398949662836642e-06, + "loss": 0.0122, + "step": 85160 + }, + { + "epoch": 24.78033168460867, + "grad_norm": 0.09835779666900635, + "learning_rate": 6.395563380665818e-06, + "loss": 0.01, + "step": 85170 + }, + { + "epoch": 24.783241198719814, + "grad_norm": 0.150889590382576, + "learning_rate": 6.392177751945165e-06, + "loss": 0.0111, + "step": 85180 + }, + { + "epoch": 24.786150712830956, + "grad_norm": 0.08370453119277954, + "learning_rate": 6.388792776931806e-06, + "loss": 0.0129, + "step": 85190 + }, + { + "epoch": 24.7890602269421, + "grad_norm": 0.09563086181879044, + "learning_rate": 6.385408455882811e-06, + "loss": 0.0122, + "step": 85200 + }, + { + "epoch": 24.791969741053244, + "grad_norm": 0.12479813396930695, + "learning_rate": 6.382024789055194e-06, + "loss": 0.0119, + "step": 85210 + }, + { + "epoch": 24.79487925516439, + "grad_norm": 0.10720158368349075, + "learning_rate": 6.378641776705912e-06, + "loss": 0.012, + "step": 85220 + }, + { + "epoch": 24.79778876927553, + "grad_norm": 0.07506361603736877, + "learning_rate": 6.375259419091894e-06, + "loss": 0.0107, + "step": 85230 + }, + { + "epoch": 24.800698283386673, + "grad_norm": 0.0850244089961052, + "learning_rate": 6.371877716470009e-06, + "loss": 0.0114, + "step": 85240 + }, + { + "epoch": 24.80360779749782, + "grad_norm": 0.13011720776557922, + "learning_rate": 6.368496669097066e-06, + "loss": 0.0131, + "step": 85250 + }, + { + "epoch": 24.80651731160896, + "grad_norm": 0.14695890247821808, + "learning_rate": 6.365116277229836e-06, + "loss": 0.0127, + "step": 85260 + }, + { + "epoch": 24.809426825720106, + "grad_norm": 0.10990049690008163, + "learning_rate": 6.361736541125047e-06, + "loss": 0.0136, + "step": 85270 + }, + { + "epoch": 24.812336339831248, + "grad_norm": 0.10284925252199173, + "learning_rate": 6.35835746103936e-06, + "loss": 0.0126, + "step": 85280 + }, + { + "epoch": 24.815245853942393, + "grad_norm": 0.10569377988576889, + "learning_rate": 6.354979037229389e-06, + "loss": 0.0125, + "step": 85290 + }, + { + "epoch": 24.818155368053535, + "grad_norm": 0.12465944141149521, + "learning_rate": 6.3516012699517065e-06, + "loss": 0.0124, + "step": 85300 + }, + { + "epoch": 24.821064882164677, + "grad_norm": 0.09027573466300964, + "learning_rate": 6.348224159462839e-06, + "loss": 0.0143, + "step": 85310 + }, + { + "epoch": 24.823974396275823, + "grad_norm": 0.1031322032213211, + "learning_rate": 6.3448477060192444e-06, + "loss": 0.0134, + "step": 85320 + }, + { + "epoch": 24.826883910386965, + "grad_norm": 0.10984901338815689, + "learning_rate": 6.341471909877353e-06, + "loss": 0.0127, + "step": 85330 + }, + { + "epoch": 24.82979342449811, + "grad_norm": 0.1103210374712944, + "learning_rate": 6.338096771293524e-06, + "loss": 0.0134, + "step": 85340 + }, + { + "epoch": 24.832702938609252, + "grad_norm": 0.10986834019422531, + "learning_rate": 6.334722290524081e-06, + "loss": 0.0117, + "step": 85350 + }, + { + "epoch": 24.835612452720397, + "grad_norm": 0.07670003175735474, + "learning_rate": 6.331348467825299e-06, + "loss": 0.0122, + "step": 85360 + }, + { + "epoch": 24.83852196683154, + "grad_norm": 0.11833784729242325, + "learning_rate": 6.327975303453388e-06, + "loss": 0.0125, + "step": 85370 + }, + { + "epoch": 24.84143148094268, + "grad_norm": 0.09652166813611984, + "learning_rate": 6.324602797664527e-06, + "loss": 0.0112, + "step": 85380 + }, + { + "epoch": 24.844340995053827, + "grad_norm": 0.10168454051017761, + "learning_rate": 6.321230950714828e-06, + "loss": 0.0124, + "step": 85390 + }, + { + "epoch": 24.84725050916497, + "grad_norm": 0.09591955691576004, + "learning_rate": 6.3178597628603666e-06, + "loss": 0.0102, + "step": 85400 + }, + { + "epoch": 24.850160023276114, + "grad_norm": 0.09037522226572037, + "learning_rate": 6.314489234357154e-06, + "loss": 0.0127, + "step": 85410 + }, + { + "epoch": 24.853069537387256, + "grad_norm": 0.07726765424013138, + "learning_rate": 6.311119365461163e-06, + "loss": 0.0115, + "step": 85420 + }, + { + "epoch": 24.8559790514984, + "grad_norm": 0.1147426962852478, + "learning_rate": 6.307750156428322e-06, + "loss": 0.0114, + "step": 85430 + }, + { + "epoch": 24.858888565609544, + "grad_norm": 0.1611948311328888, + "learning_rate": 6.304381607514493e-06, + "loss": 0.0122, + "step": 85440 + }, + { + "epoch": 24.861798079720685, + "grad_norm": 0.12793035805225372, + "learning_rate": 6.301013718975483e-06, + "loss": 0.0132, + "step": 85450 + }, + { + "epoch": 24.86470759383183, + "grad_norm": 0.16777218878269196, + "learning_rate": 6.297646491067085e-06, + "loss": 0.0123, + "step": 85460 + }, + { + "epoch": 24.867617107942973, + "grad_norm": 0.11690624058246613, + "learning_rate": 6.294279924045003e-06, + "loss": 0.0122, + "step": 85470 + }, + { + "epoch": 24.87052662205412, + "grad_norm": 0.10180002450942993, + "learning_rate": 6.290914018164905e-06, + "loss": 0.012, + "step": 85480 + }, + { + "epoch": 24.87343613616526, + "grad_norm": 0.10044217854738235, + "learning_rate": 6.287548773682411e-06, + "loss": 0.0137, + "step": 85490 + }, + { + "epoch": 24.876345650276402, + "grad_norm": 0.20322756469249725, + "learning_rate": 6.2841841908530955e-06, + "loss": 0.012, + "step": 85500 + }, + { + "epoch": 24.879255164387548, + "grad_norm": 0.058610253036022186, + "learning_rate": 6.280820269932472e-06, + "loss": 0.0106, + "step": 85510 + }, + { + "epoch": 24.88216467849869, + "grad_norm": 0.10222066938877106, + "learning_rate": 6.277457011176003e-06, + "loss": 0.0117, + "step": 85520 + }, + { + "epoch": 24.885074192609835, + "grad_norm": 0.09049922227859497, + "learning_rate": 6.274094414839109e-06, + "loss": 0.0107, + "step": 85530 + }, + { + "epoch": 24.887983706720977, + "grad_norm": 0.0967082679271698, + "learning_rate": 6.270732481177165e-06, + "loss": 0.0133, + "step": 85540 + }, + { + "epoch": 24.890893220832123, + "grad_norm": 0.0962807759642601, + "learning_rate": 6.267371210445473e-06, + "loss": 0.0105, + "step": 85550 + }, + { + "epoch": 24.893802734943264, + "grad_norm": 0.11145221441984177, + "learning_rate": 6.26401060289931e-06, + "loss": 0.0123, + "step": 85560 + }, + { + "epoch": 24.896712249054406, + "grad_norm": 0.10226771980524063, + "learning_rate": 6.260650658793892e-06, + "loss": 0.0125, + "step": 85570 + }, + { + "epoch": 24.899621763165552, + "grad_norm": 0.15415678918361664, + "learning_rate": 6.2572913783843845e-06, + "loss": 0.0136, + "step": 85580 + }, + { + "epoch": 24.902531277276694, + "grad_norm": 0.11573503911495209, + "learning_rate": 6.253932761925893e-06, + "loss": 0.0116, + "step": 85590 + }, + { + "epoch": 24.90544079138784, + "grad_norm": 0.1452203392982483, + "learning_rate": 6.250574809673489e-06, + "loss": 0.0102, + "step": 85600 + }, + { + "epoch": 24.90835030549898, + "grad_norm": 0.14916515350341797, + "learning_rate": 6.247217521882193e-06, + "loss": 0.0126, + "step": 85610 + }, + { + "epoch": 24.911259819610127, + "grad_norm": 0.10076643526554108, + "learning_rate": 6.243860898806956e-06, + "loss": 0.013, + "step": 85620 + }, + { + "epoch": 24.91416933372127, + "grad_norm": 0.10956788063049316, + "learning_rate": 6.240504940702706e-06, + "loss": 0.0124, + "step": 85630 + }, + { + "epoch": 24.91707884783241, + "grad_norm": 0.10100875794887543, + "learning_rate": 6.23714964782429e-06, + "loss": 0.0117, + "step": 85640 + }, + { + "epoch": 24.919988361943556, + "grad_norm": 0.15398548543453217, + "learning_rate": 6.2337950204265316e-06, + "loss": 0.0131, + "step": 85650 + }, + { + "epoch": 24.922897876054698, + "grad_norm": 0.10184542834758759, + "learning_rate": 6.230441058764193e-06, + "loss": 0.0134, + "step": 85660 + }, + { + "epoch": 24.925807390165843, + "grad_norm": 0.13151587545871735, + "learning_rate": 6.227087763091976e-06, + "loss": 0.0102, + "step": 85670 + }, + { + "epoch": 24.928716904276985, + "grad_norm": 0.10899757593870163, + "learning_rate": 6.223735133664553e-06, + "loss": 0.012, + "step": 85680 + }, + { + "epoch": 24.93162641838813, + "grad_norm": 0.18516609072685242, + "learning_rate": 6.220383170736524e-06, + "loss": 0.0098, + "step": 85690 + }, + { + "epoch": 24.934535932499273, + "grad_norm": 0.09418153017759323, + "learning_rate": 6.217031874562458e-06, + "loss": 0.0122, + "step": 85700 + }, + { + "epoch": 24.937445446610415, + "grad_norm": 0.09670817852020264, + "learning_rate": 6.213681245396853e-06, + "loss": 0.0114, + "step": 85710 + }, + { + "epoch": 24.94035496072156, + "grad_norm": 0.13291838765144348, + "learning_rate": 6.210331283494173e-06, + "loss": 0.011, + "step": 85720 + }, + { + "epoch": 24.943264474832702, + "grad_norm": 0.11332960426807404, + "learning_rate": 6.206981989108829e-06, + "loss": 0.0112, + "step": 85730 + }, + { + "epoch": 24.946173988943848, + "grad_norm": 0.09275561571121216, + "learning_rate": 6.203633362495175e-06, + "loss": 0.0126, + "step": 85740 + }, + { + "epoch": 24.94908350305499, + "grad_norm": 0.1205025166273117, + "learning_rate": 6.2002854039075105e-06, + "loss": 0.011, + "step": 85750 + }, + { + "epoch": 24.95199301716613, + "grad_norm": 0.13624876737594604, + "learning_rate": 6.196938113600098e-06, + "loss": 0.013, + "step": 85760 + }, + { + "epoch": 24.954902531277277, + "grad_norm": 0.13278517127037048, + "learning_rate": 6.1935914918271455e-06, + "loss": 0.0118, + "step": 85770 + }, + { + "epoch": 24.95781204538842, + "grad_norm": 0.09846209734678268, + "learning_rate": 6.190245538842794e-06, + "loss": 0.0105, + "step": 85780 + }, + { + "epoch": 24.960721559499564, + "grad_norm": 0.11793617159128189, + "learning_rate": 6.186900254901158e-06, + "loss": 0.013, + "step": 85790 + }, + { + "epoch": 24.963631073610706, + "grad_norm": 0.11861302703619003, + "learning_rate": 6.183555640256289e-06, + "loss": 0.0136, + "step": 85800 + }, + { + "epoch": 24.966540587721852, + "grad_norm": 0.09886655956506729, + "learning_rate": 6.180211695162188e-06, + "loss": 0.0117, + "step": 85810 + }, + { + "epoch": 24.969450101832994, + "grad_norm": 0.12021879851818085, + "learning_rate": 6.1768684198727955e-06, + "loss": 0.0154, + "step": 85820 + }, + { + "epoch": 24.972359615944136, + "grad_norm": 0.11797337234020233, + "learning_rate": 6.173525814642021e-06, + "loss": 0.0131, + "step": 85830 + }, + { + "epoch": 24.97526913005528, + "grad_norm": 0.11607497930526733, + "learning_rate": 6.1701838797237146e-06, + "loss": 0.013, + "step": 85840 + }, + { + "epoch": 24.978178644166423, + "grad_norm": 0.13021092116832733, + "learning_rate": 6.166842615371672e-06, + "loss": 0.0109, + "step": 85850 + }, + { + "epoch": 24.98108815827757, + "grad_norm": 0.13061492145061493, + "learning_rate": 6.163502021839628e-06, + "loss": 0.0099, + "step": 85860 + }, + { + "epoch": 24.98399767238871, + "grad_norm": 0.13041238486766815, + "learning_rate": 6.160162099381301e-06, + "loss": 0.0107, + "step": 85870 + }, + { + "epoch": 24.986907186499856, + "grad_norm": 0.06641504168510437, + "learning_rate": 6.156822848250324e-06, + "loss": 0.0109, + "step": 85880 + }, + { + "epoch": 24.989816700610998, + "grad_norm": 0.08063560724258423, + "learning_rate": 6.153484268700286e-06, + "loss": 0.0114, + "step": 85890 + }, + { + "epoch": 24.99272621472214, + "grad_norm": 0.09768492728471756, + "learning_rate": 6.150146360984738e-06, + "loss": 0.0119, + "step": 85900 + }, + { + "epoch": 24.995635728833285, + "grad_norm": 0.09927854686975479, + "learning_rate": 6.146809125357173e-06, + "loss": 0.0119, + "step": 85910 + }, + { + "epoch": 24.998545242944427, + "grad_norm": 0.0972532257437706, + "learning_rate": 6.1434725620710245e-06, + "loss": 0.0123, + "step": 85920 + }, + { + "epoch": 25.001454757055573, + "grad_norm": 0.11746270954608917, + "learning_rate": 6.140136671379692e-06, + "loss": 0.0107, + "step": 85930 + }, + { + "epoch": 25.004364271166715, + "grad_norm": 0.1110803559422493, + "learning_rate": 6.136801453536503e-06, + "loss": 0.0127, + "step": 85940 + }, + { + "epoch": 25.00727378527786, + "grad_norm": 0.08417966961860657, + "learning_rate": 6.133466908794758e-06, + "loss": 0.012, + "step": 85950 + }, + { + "epoch": 25.010183299389002, + "grad_norm": 0.10145967453718185, + "learning_rate": 6.130133037407683e-06, + "loss": 0.0129, + "step": 85960 + }, + { + "epoch": 25.013092813500144, + "grad_norm": 0.14896011352539062, + "learning_rate": 6.126799839628466e-06, + "loss": 0.0113, + "step": 85970 + }, + { + "epoch": 25.01600232761129, + "grad_norm": 0.12887533009052277, + "learning_rate": 6.123467315710249e-06, + "loss": 0.0119, + "step": 85980 + }, + { + "epoch": 25.01891184172243, + "grad_norm": 0.17322488129138947, + "learning_rate": 6.1201354659061045e-06, + "loss": 0.012, + "step": 85990 + }, + { + "epoch": 25.021821355833577, + "grad_norm": 0.19435974955558777, + "learning_rate": 6.116804290469075e-06, + "loss": 0.0142, + "step": 86000 + }, + { + "epoch": 25.02473086994472, + "grad_norm": 0.11961536854505539, + "learning_rate": 6.1134737896521295e-06, + "loss": 0.0125, + "step": 86010 + }, + { + "epoch": 25.027640384055864, + "grad_norm": 0.0866071805357933, + "learning_rate": 6.110143963708203e-06, + "loss": 0.0113, + "step": 86020 + }, + { + "epoch": 25.030549898167006, + "grad_norm": 0.0879877582192421, + "learning_rate": 6.106814812890182e-06, + "loss": 0.0104, + "step": 86030 + }, + { + "epoch": 25.033459412278148, + "grad_norm": 0.07898099720478058, + "learning_rate": 6.103486337450884e-06, + "loss": 0.0115, + "step": 86040 + }, + { + "epoch": 25.036368926389294, + "grad_norm": 0.12815001606941223, + "learning_rate": 6.100158537643081e-06, + "loss": 0.0116, + "step": 86050 + }, + { + "epoch": 25.039278440500436, + "grad_norm": 0.11045181006193161, + "learning_rate": 6.0968314137195055e-06, + "loss": 0.0125, + "step": 86060 + }, + { + "epoch": 25.04218795461158, + "grad_norm": 0.08950744569301605, + "learning_rate": 6.093504965932832e-06, + "loss": 0.0113, + "step": 86070 + }, + { + "epoch": 25.045097468722723, + "grad_norm": 0.09602786600589752, + "learning_rate": 6.090179194535674e-06, + "loss": 0.0125, + "step": 86080 + }, + { + "epoch": 25.04800698283387, + "grad_norm": 0.14716146886348724, + "learning_rate": 6.086854099780607e-06, + "loss": 0.0139, + "step": 86090 + }, + { + "epoch": 25.05091649694501, + "grad_norm": 0.08243688195943832, + "learning_rate": 6.083529681920155e-06, + "loss": 0.0141, + "step": 86100 + }, + { + "epoch": 25.053826011056152, + "grad_norm": 0.12707296013832092, + "learning_rate": 6.080205941206778e-06, + "loss": 0.0138, + "step": 86110 + }, + { + "epoch": 25.056735525167298, + "grad_norm": 0.09195158630609512, + "learning_rate": 6.076882877892891e-06, + "loss": 0.0125, + "step": 86120 + }, + { + "epoch": 25.05964503927844, + "grad_norm": 0.09495742619037628, + "learning_rate": 6.0735604922308605e-06, + "loss": 0.0131, + "step": 86130 + }, + { + "epoch": 25.062554553389585, + "grad_norm": 0.1219615563750267, + "learning_rate": 6.070238784473007e-06, + "loss": 0.0108, + "step": 86140 + }, + { + "epoch": 25.065464067500727, + "grad_norm": 0.11711146682500839, + "learning_rate": 6.066917754871587e-06, + "loss": 0.0103, + "step": 86150 + }, + { + "epoch": 25.06837358161187, + "grad_norm": 0.11943256109952927, + "learning_rate": 6.063597403678799e-06, + "loss": 0.0123, + "step": 86160 + }, + { + "epoch": 25.071283095723015, + "grad_norm": 0.11109407246112823, + "learning_rate": 6.060277731146825e-06, + "loss": 0.0133, + "step": 86170 + }, + { + "epoch": 25.074192609834157, + "grad_norm": 0.11558933556079865, + "learning_rate": 6.056958737527759e-06, + "loss": 0.012, + "step": 86180 + }, + { + "epoch": 25.077102123945302, + "grad_norm": 0.11845910549163818, + "learning_rate": 6.053640423073652e-06, + "loss": 0.0098, + "step": 86190 + }, + { + "epoch": 25.080011638056444, + "grad_norm": 0.13331936299800873, + "learning_rate": 6.050322788036515e-06, + "loss": 0.0109, + "step": 86200 + }, + { + "epoch": 25.08292115216759, + "grad_norm": 0.10883145779371262, + "learning_rate": 6.0470058326683055e-06, + "loss": 0.0141, + "step": 86210 + }, + { + "epoch": 25.08583066627873, + "grad_norm": 0.12354930490255356, + "learning_rate": 6.043689557220918e-06, + "loss": 0.0095, + "step": 86220 + }, + { + "epoch": 25.088740180389873, + "grad_norm": 0.12159726768732071, + "learning_rate": 6.040373961946196e-06, + "loss": 0.0125, + "step": 86230 + }, + { + "epoch": 25.09164969450102, + "grad_norm": 0.12891516089439392, + "learning_rate": 6.037059047095944e-06, + "loss": 0.0128, + "step": 86240 + }, + { + "epoch": 25.09455920861216, + "grad_norm": 0.08604128658771515, + "learning_rate": 6.033744812921911e-06, + "loss": 0.0118, + "step": 86250 + }, + { + "epoch": 25.097468722723306, + "grad_norm": 0.09768880158662796, + "learning_rate": 6.030431259675784e-06, + "loss": 0.0119, + "step": 86260 + }, + { + "epoch": 25.100378236834448, + "grad_norm": 0.09811443835496902, + "learning_rate": 6.027118387609208e-06, + "loss": 0.0125, + "step": 86270 + }, + { + "epoch": 25.103287750945594, + "grad_norm": 0.07936377078294754, + "learning_rate": 6.02380619697378e-06, + "loss": 0.0125, + "step": 86280 + }, + { + "epoch": 25.106197265056736, + "grad_norm": 0.08743051439523697, + "learning_rate": 6.020494688021028e-06, + "loss": 0.0107, + "step": 86290 + }, + { + "epoch": 25.109106779167877, + "grad_norm": 0.14093996584415436, + "learning_rate": 6.017183861002452e-06, + "loss": 0.0114, + "step": 86300 + }, + { + "epoch": 25.112016293279023, + "grad_norm": 0.11149803549051285, + "learning_rate": 6.013873716169473e-06, + "loss": 0.0134, + "step": 86310 + }, + { + "epoch": 25.114925807390165, + "grad_norm": 0.10292930155992508, + "learning_rate": 6.010564253773486e-06, + "loss": 0.0102, + "step": 86320 + }, + { + "epoch": 25.11783532150131, + "grad_norm": 0.11240074038505554, + "learning_rate": 6.0072554740658154e-06, + "loss": 0.0131, + "step": 86330 + }, + { + "epoch": 25.120744835612452, + "grad_norm": 0.14069773256778717, + "learning_rate": 6.003947377297751e-06, + "loss": 0.0121, + "step": 86340 + }, + { + "epoch": 25.123654349723598, + "grad_norm": 0.13305328786373138, + "learning_rate": 6.000639963720507e-06, + "loss": 0.0126, + "step": 86350 + }, + { + "epoch": 25.12656386383474, + "grad_norm": 0.1065889298915863, + "learning_rate": 5.997333233585268e-06, + "loss": 0.0134, + "step": 86360 + }, + { + "epoch": 25.12947337794588, + "grad_norm": 0.0846293717622757, + "learning_rate": 5.994027187143163e-06, + "loss": 0.0104, + "step": 86370 + }, + { + "epoch": 25.132382892057027, + "grad_norm": 0.15310800075531006, + "learning_rate": 5.990721824645254e-06, + "loss": 0.0126, + "step": 86380 + }, + { + "epoch": 25.13529240616817, + "grad_norm": 0.10960552841424942, + "learning_rate": 5.987417146342565e-06, + "loss": 0.0114, + "step": 86390 + }, + { + "epoch": 25.138201920279315, + "grad_norm": 0.0988236665725708, + "learning_rate": 5.9841131524860735e-06, + "loss": 0.0118, + "step": 86400 + }, + { + "epoch": 25.141111434390456, + "grad_norm": 0.1094384491443634, + "learning_rate": 5.980809843326685e-06, + "loss": 0.0124, + "step": 86410 + }, + { + "epoch": 25.144020948501602, + "grad_norm": 0.11255045235157013, + "learning_rate": 5.977507219115265e-06, + "loss": 0.012, + "step": 86420 + }, + { + "epoch": 25.146930462612744, + "grad_norm": 0.11158303171396255, + "learning_rate": 5.9742052801026265e-06, + "loss": 0.0126, + "step": 86430 + }, + { + "epoch": 25.149839976723886, + "grad_norm": 0.08343951404094696, + "learning_rate": 5.970904026539538e-06, + "loss": 0.0136, + "step": 86440 + }, + { + "epoch": 25.15274949083503, + "grad_norm": 0.12121092528104782, + "learning_rate": 5.967603458676703e-06, + "loss": 0.0116, + "step": 86450 + }, + { + "epoch": 25.155659004946173, + "grad_norm": 0.08739136904478073, + "learning_rate": 5.964303576764764e-06, + "loss": 0.0104, + "step": 86460 + }, + { + "epoch": 25.15856851905732, + "grad_norm": 0.10387146472930908, + "learning_rate": 5.961004381054351e-06, + "loss": 0.0146, + "step": 86470 + }, + { + "epoch": 25.16147803316846, + "grad_norm": 0.10584622621536255, + "learning_rate": 5.957705871796001e-06, + "loss": 0.0148, + "step": 86480 + }, + { + "epoch": 25.164387547279603, + "grad_norm": 0.11739082634449005, + "learning_rate": 5.954408049240211e-06, + "loss": 0.0126, + "step": 86490 + }, + { + "epoch": 25.167297061390748, + "grad_norm": 0.09067046642303467, + "learning_rate": 5.951110913637435e-06, + "loss": 0.0107, + "step": 86500 + }, + { + "epoch": 25.17020657550189, + "grad_norm": 0.101159006357193, + "learning_rate": 5.9478144652380725e-06, + "loss": 0.0128, + "step": 86510 + }, + { + "epoch": 25.173116089613035, + "grad_norm": 0.09772844612598419, + "learning_rate": 5.9445187042924615e-06, + "loss": 0.0123, + "step": 86520 + }, + { + "epoch": 25.176025603724177, + "grad_norm": 0.14358943700790405, + "learning_rate": 5.9412236310508896e-06, + "loss": 0.0137, + "step": 86530 + }, + { + "epoch": 25.178935117835323, + "grad_norm": 0.10577183216810226, + "learning_rate": 5.937929245763599e-06, + "loss": 0.0128, + "step": 86540 + }, + { + "epoch": 25.181844631946465, + "grad_norm": 0.13220950961112976, + "learning_rate": 5.934635548680784e-06, + "loss": 0.0134, + "step": 86550 + }, + { + "epoch": 25.184754146057607, + "grad_norm": 0.07032065838575363, + "learning_rate": 5.931342540052567e-06, + "loss": 0.012, + "step": 86560 + }, + { + "epoch": 25.187663660168752, + "grad_norm": 0.13498838245868683, + "learning_rate": 5.928050220129037e-06, + "loss": 0.0123, + "step": 86570 + }, + { + "epoch": 25.190573174279894, + "grad_norm": 0.12329194694757462, + "learning_rate": 5.9247585891602265e-06, + "loss": 0.0125, + "step": 86580 + }, + { + "epoch": 25.19348268839104, + "grad_norm": 0.11143108457326889, + "learning_rate": 5.921467647396112e-06, + "loss": 0.0127, + "step": 86590 + }, + { + "epoch": 25.19639220250218, + "grad_norm": 0.10464686155319214, + "learning_rate": 5.9181773950866114e-06, + "loss": 0.0117, + "step": 86600 + }, + { + "epoch": 25.199301716613327, + "grad_norm": 0.07959343492984772, + "learning_rate": 5.9148878324816016e-06, + "loss": 0.0129, + "step": 86610 + }, + { + "epoch": 25.20221123072447, + "grad_norm": 0.1128874197602272, + "learning_rate": 5.91159895983091e-06, + "loss": 0.0123, + "step": 86620 + }, + { + "epoch": 25.20512074483561, + "grad_norm": 0.11910408735275269, + "learning_rate": 5.908310777384292e-06, + "loss": 0.0114, + "step": 86630 + }, + { + "epoch": 25.208030258946756, + "grad_norm": 0.10409354418516159, + "learning_rate": 5.905023285391479e-06, + "loss": 0.0112, + "step": 86640 + }, + { + "epoch": 25.2109397730579, + "grad_norm": 0.12335251271724701, + "learning_rate": 5.901736484102119e-06, + "loss": 0.012, + "step": 86650 + }, + { + "epoch": 25.213849287169044, + "grad_norm": 0.11148657649755478, + "learning_rate": 5.898450373765829e-06, + "loss": 0.012, + "step": 86660 + }, + { + "epoch": 25.216758801280186, + "grad_norm": 0.10629821568727493, + "learning_rate": 5.895164954632176e-06, + "loss": 0.0096, + "step": 86670 + }, + { + "epoch": 25.21966831539133, + "grad_norm": 0.12796781957149506, + "learning_rate": 5.891880226950651e-06, + "loss": 0.0116, + "step": 86680 + }, + { + "epoch": 25.222577829502473, + "grad_norm": 0.09295260906219482, + "learning_rate": 5.88859619097072e-06, + "loss": 0.0106, + "step": 86690 + }, + { + "epoch": 25.225487343613615, + "grad_norm": 0.1172393187880516, + "learning_rate": 5.885312846941772e-06, + "loss": 0.0122, + "step": 86700 + }, + { + "epoch": 25.22839685772476, + "grad_norm": 0.11674809455871582, + "learning_rate": 5.882030195113166e-06, + "loss": 0.0116, + "step": 86710 + }, + { + "epoch": 25.231306371835903, + "grad_norm": 0.0941578820347786, + "learning_rate": 5.87874823573419e-06, + "loss": 0.0109, + "step": 86720 + }, + { + "epoch": 25.234215885947048, + "grad_norm": 0.10578205436468124, + "learning_rate": 5.8754669690540905e-06, + "loss": 0.0127, + "step": 86730 + }, + { + "epoch": 25.23712540005819, + "grad_norm": 0.0962415337562561, + "learning_rate": 5.872186395322063e-06, + "loss": 0.011, + "step": 86740 + }, + { + "epoch": 25.240034914169335, + "grad_norm": 0.11222007870674133, + "learning_rate": 5.8689065147872404e-06, + "loss": 0.0113, + "step": 86750 + }, + { + "epoch": 25.242944428280477, + "grad_norm": 0.10162713378667831, + "learning_rate": 5.865627327698698e-06, + "loss": 0.0112, + "step": 86760 + }, + { + "epoch": 25.24585394239162, + "grad_norm": 0.1168115884065628, + "learning_rate": 5.862348834305489e-06, + "loss": 0.0131, + "step": 86770 + }, + { + "epoch": 25.248763456502765, + "grad_norm": 0.07431214302778244, + "learning_rate": 5.859071034856582e-06, + "loss": 0.0133, + "step": 86780 + }, + { + "epoch": 25.251672970613907, + "grad_norm": 0.1111876592040062, + "learning_rate": 5.8557939296009e-06, + "loss": 0.0121, + "step": 86790 + }, + { + "epoch": 25.254582484725052, + "grad_norm": 0.09611357748508453, + "learning_rate": 5.852517518787323e-06, + "loss": 0.0113, + "step": 86800 + }, + { + "epoch": 25.257491998836194, + "grad_norm": 0.09273262321949005, + "learning_rate": 5.8492418026646766e-06, + "loss": 0.0117, + "step": 86810 + }, + { + "epoch": 25.26040151294734, + "grad_norm": 0.1644279956817627, + "learning_rate": 5.845966781481725e-06, + "loss": 0.0132, + "step": 86820 + }, + { + "epoch": 25.26331102705848, + "grad_norm": 0.1441599726676941, + "learning_rate": 5.842692455487181e-06, + "loss": 0.0136, + "step": 86830 + }, + { + "epoch": 25.266220541169623, + "grad_norm": 0.2767721116542816, + "learning_rate": 5.839418824929713e-06, + "loss": 0.012, + "step": 86840 + }, + { + "epoch": 25.26913005528077, + "grad_norm": 0.1296311616897583, + "learning_rate": 5.836145890057933e-06, + "loss": 0.0134, + "step": 86850 + }, + { + "epoch": 25.27203956939191, + "grad_norm": 0.0798201784491539, + "learning_rate": 5.832873651120394e-06, + "loss": 0.0106, + "step": 86860 + }, + { + "epoch": 25.274949083503056, + "grad_norm": 0.1196063980460167, + "learning_rate": 5.829602108365602e-06, + "loss": 0.0127, + "step": 86870 + }, + { + "epoch": 25.2778585976142, + "grad_norm": 0.08389527350664139, + "learning_rate": 5.826331262042015e-06, + "loss": 0.0107, + "step": 86880 + }, + { + "epoch": 25.28076811172534, + "grad_norm": 0.08662257343530655, + "learning_rate": 5.823061112398031e-06, + "loss": 0.0115, + "step": 86890 + }, + { + "epoch": 25.283677625836486, + "grad_norm": 0.08979322016239166, + "learning_rate": 5.8197916596819844e-06, + "loss": 0.0109, + "step": 86900 + }, + { + "epoch": 25.286587139947628, + "grad_norm": 0.1421636939048767, + "learning_rate": 5.816522904142179e-06, + "loss": 0.0138, + "step": 86910 + }, + { + "epoch": 25.289496654058773, + "grad_norm": 0.13639117777347565, + "learning_rate": 5.8132548460268575e-06, + "loss": 0.0125, + "step": 86920 + }, + { + "epoch": 25.292406168169915, + "grad_norm": 0.1142483577132225, + "learning_rate": 5.8099874855841974e-06, + "loss": 0.0115, + "step": 86930 + }, + { + "epoch": 25.29531568228106, + "grad_norm": 0.09981069713830948, + "learning_rate": 5.8067208230623455e-06, + "loss": 0.0133, + "step": 86940 + }, + { + "epoch": 25.298225196392202, + "grad_norm": 0.17936305701732635, + "learning_rate": 5.803454858709371e-06, + "loss": 0.0111, + "step": 86950 + }, + { + "epoch": 25.301134710503344, + "grad_norm": 0.12150456011295319, + "learning_rate": 5.8001895927733135e-06, + "loss": 0.0121, + "step": 86960 + }, + { + "epoch": 25.30404422461449, + "grad_norm": 0.14500518143177032, + "learning_rate": 5.796925025502138e-06, + "loss": 0.0149, + "step": 86970 + }, + { + "epoch": 25.306953738725632, + "grad_norm": 0.12600308656692505, + "learning_rate": 5.793661157143772e-06, + "loss": 0.0113, + "step": 86980 + }, + { + "epoch": 25.309863252836777, + "grad_norm": 0.12201393395662308, + "learning_rate": 5.790397987946088e-06, + "loss": 0.0122, + "step": 86990 + }, + { + "epoch": 25.31277276694792, + "grad_norm": 0.08881831914186478, + "learning_rate": 5.787135518156895e-06, + "loss": 0.0143, + "step": 87000 + }, + { + "epoch": 25.315682281059065, + "grad_norm": 0.10068480670452118, + "learning_rate": 5.783873748023966e-06, + "loss": 0.013, + "step": 87010 + }, + { + "epoch": 25.318591795170207, + "grad_norm": 0.08728103339672089, + "learning_rate": 5.7806126777949975e-06, + "loss": 0.0112, + "step": 87020 + }, + { + "epoch": 25.32150130928135, + "grad_norm": 0.10652013123035431, + "learning_rate": 5.7773523077176544e-06, + "loss": 0.0114, + "step": 87030 + }, + { + "epoch": 25.324410823392494, + "grad_norm": 0.12219797819852829, + "learning_rate": 5.774092638039543e-06, + "loss": 0.0132, + "step": 87040 + }, + { + "epoch": 25.327320337503636, + "grad_norm": 0.08915212750434875, + "learning_rate": 5.770833669008213e-06, + "loss": 0.0102, + "step": 87050 + }, + { + "epoch": 25.33022985161478, + "grad_norm": 0.12581028044223785, + "learning_rate": 5.7675754008711515e-06, + "loss": 0.0138, + "step": 87060 + }, + { + "epoch": 25.333139365725923, + "grad_norm": 0.12684214115142822, + "learning_rate": 5.76431783387581e-06, + "loss": 0.0136, + "step": 87070 + }, + { + "epoch": 25.33604887983707, + "grad_norm": 0.08699925988912582, + "learning_rate": 5.761060968269583e-06, + "loss": 0.0126, + "step": 87080 + }, + { + "epoch": 25.33895839394821, + "grad_norm": 0.11143616586923599, + "learning_rate": 5.757804804299801e-06, + "loss": 0.0092, + "step": 87090 + }, + { + "epoch": 25.341867908059353, + "grad_norm": 0.13559377193450928, + "learning_rate": 5.754549342213749e-06, + "loss": 0.0154, + "step": 87100 + }, + { + "epoch": 25.3447774221705, + "grad_norm": 0.0945858359336853, + "learning_rate": 5.751294582258665e-06, + "loss": 0.0124, + "step": 87110 + }, + { + "epoch": 25.34768693628164, + "grad_norm": 0.09698645025491714, + "learning_rate": 5.748040524681722e-06, + "loss": 0.0134, + "step": 87120 + }, + { + "epoch": 25.350596450392786, + "grad_norm": 0.1870017945766449, + "learning_rate": 5.744787169730037e-06, + "loss": 0.0111, + "step": 87130 + }, + { + "epoch": 25.353505964503928, + "grad_norm": 0.13926786184310913, + "learning_rate": 5.7415345176506905e-06, + "loss": 0.0107, + "step": 87140 + }, + { + "epoch": 25.35641547861507, + "grad_norm": 0.07746720314025879, + "learning_rate": 5.738282568690699e-06, + "loss": 0.0119, + "step": 87150 + }, + { + "epoch": 25.359324992726215, + "grad_norm": 0.11385714262723923, + "learning_rate": 5.7350313230970255e-06, + "loss": 0.0105, + "step": 87160 + }, + { + "epoch": 25.362234506837357, + "grad_norm": 0.07208363711833954, + "learning_rate": 5.731780781116569e-06, + "loss": 0.0111, + "step": 87170 + }, + { + "epoch": 25.365144020948502, + "grad_norm": 0.11877432465553284, + "learning_rate": 5.728530942996207e-06, + "loss": 0.0131, + "step": 87180 + }, + { + "epoch": 25.368053535059644, + "grad_norm": 0.10301011055707932, + "learning_rate": 5.725281808982735e-06, + "loss": 0.0137, + "step": 87190 + }, + { + "epoch": 25.37096304917079, + "grad_norm": 0.08781221508979797, + "learning_rate": 5.722033379322895e-06, + "loss": 0.0118, + "step": 87200 + }, + { + "epoch": 25.37387256328193, + "grad_norm": 0.26018327474594116, + "learning_rate": 5.718785654263392e-06, + "loss": 0.0123, + "step": 87210 + }, + { + "epoch": 25.376782077393074, + "grad_norm": 0.11429275572299957, + "learning_rate": 5.715538634050872e-06, + "loss": 0.0117, + "step": 87220 + }, + { + "epoch": 25.37969159150422, + "grad_norm": 0.16038130223751068, + "learning_rate": 5.712292318931916e-06, + "loss": 0.0125, + "step": 87230 + }, + { + "epoch": 25.38260110561536, + "grad_norm": 0.11496246606111526, + "learning_rate": 5.70904670915307e-06, + "loss": 0.0124, + "step": 87240 + }, + { + "epoch": 25.385510619726507, + "grad_norm": 0.09996509552001953, + "learning_rate": 5.705801804960805e-06, + "loss": 0.0128, + "step": 87250 + }, + { + "epoch": 25.38842013383765, + "grad_norm": 0.11457006633281708, + "learning_rate": 5.702557606601564e-06, + "loss": 0.0137, + "step": 87260 + }, + { + "epoch": 25.391329647948794, + "grad_norm": 0.12397462874650955, + "learning_rate": 5.699314114321709e-06, + "loss": 0.0136, + "step": 87270 + }, + { + "epoch": 25.394239162059936, + "grad_norm": 0.11455778777599335, + "learning_rate": 5.696071328367568e-06, + "loss": 0.0147, + "step": 87280 + }, + { + "epoch": 25.397148676171078, + "grad_norm": 0.11188450455665588, + "learning_rate": 5.692829248985414e-06, + "loss": 0.0121, + "step": 87290 + }, + { + "epoch": 25.400058190282223, + "grad_norm": 0.1751265972852707, + "learning_rate": 5.689587876421451e-06, + "loss": 0.0113, + "step": 87300 + }, + { + "epoch": 25.402967704393365, + "grad_norm": 0.09795019775629044, + "learning_rate": 5.6863472109218506e-06, + "loss": 0.0138, + "step": 87310 + }, + { + "epoch": 25.40587721850451, + "grad_norm": 0.12966565787792206, + "learning_rate": 5.683107252732709e-06, + "loss": 0.0124, + "step": 87320 + }, + { + "epoch": 25.408786732615653, + "grad_norm": 0.12447944283485413, + "learning_rate": 5.679868002100092e-06, + "loss": 0.0136, + "step": 87330 + }, + { + "epoch": 25.411696246726798, + "grad_norm": 0.0785774365067482, + "learning_rate": 5.676629459269987e-06, + "loss": 0.0104, + "step": 87340 + }, + { + "epoch": 25.41460576083794, + "grad_norm": 0.09878047555685043, + "learning_rate": 5.67339162448835e-06, + "loss": 0.0132, + "step": 87350 + }, + { + "epoch": 25.417515274949082, + "grad_norm": 0.1129290759563446, + "learning_rate": 5.670154498001064e-06, + "loss": 0.0112, + "step": 87360 + }, + { + "epoch": 25.420424789060228, + "grad_norm": 0.1084652841091156, + "learning_rate": 5.666918080053974e-06, + "loss": 0.0138, + "step": 87370 + }, + { + "epoch": 25.42333430317137, + "grad_norm": 0.10602705925703049, + "learning_rate": 5.663682370892867e-06, + "loss": 0.0112, + "step": 87380 + }, + { + "epoch": 25.426243817282515, + "grad_norm": 0.1279289871454239, + "learning_rate": 5.660447370763464e-06, + "loss": 0.0134, + "step": 87390 + }, + { + "epoch": 25.429153331393657, + "grad_norm": 0.11082940548658371, + "learning_rate": 5.657213079911447e-06, + "loss": 0.0122, + "step": 87400 + }, + { + "epoch": 25.432062845504802, + "grad_norm": 0.12318835407495499, + "learning_rate": 5.653979498582445e-06, + "loss": 0.0139, + "step": 87410 + }, + { + "epoch": 25.434972359615944, + "grad_norm": 0.10859766602516174, + "learning_rate": 5.650746627022022e-06, + "loss": 0.0114, + "step": 87420 + }, + { + "epoch": 25.437881873727086, + "grad_norm": 0.08902312070131302, + "learning_rate": 5.647514465475689e-06, + "loss": 0.0128, + "step": 87430 + }, + { + "epoch": 25.44079138783823, + "grad_norm": 0.14400836825370789, + "learning_rate": 5.6442830141889094e-06, + "loss": 0.0122, + "step": 87440 + }, + { + "epoch": 25.443700901949374, + "grad_norm": 0.14861242473125458, + "learning_rate": 5.641052273407098e-06, + "loss": 0.0124, + "step": 87450 + }, + { + "epoch": 25.44661041606052, + "grad_norm": 0.122688427567482, + "learning_rate": 5.637822243375604e-06, + "loss": 0.0159, + "step": 87460 + }, + { + "epoch": 25.44951993017166, + "grad_norm": 0.10872464627027512, + "learning_rate": 5.634592924339714e-06, + "loss": 0.0116, + "step": 87470 + }, + { + "epoch": 25.452429444282807, + "grad_norm": 0.07674426585435867, + "learning_rate": 5.631364316544696e-06, + "loss": 0.012, + "step": 87480 + }, + { + "epoch": 25.45533895839395, + "grad_norm": 0.1386343538761139, + "learning_rate": 5.628136420235731e-06, + "loss": 0.01, + "step": 87490 + }, + { + "epoch": 25.45824847250509, + "grad_norm": 0.07711838185787201, + "learning_rate": 5.62490923565795e-06, + "loss": 0.0128, + "step": 87500 + }, + { + "epoch": 25.461157986616236, + "grad_norm": 0.09883510321378708, + "learning_rate": 5.621682763056443e-06, + "loss": 0.0111, + "step": 87510 + }, + { + "epoch": 25.464067500727378, + "grad_norm": 0.13884173333644867, + "learning_rate": 5.6184570026762425e-06, + "loss": 0.0144, + "step": 87520 + }, + { + "epoch": 25.466977014838523, + "grad_norm": 0.11150261759757996, + "learning_rate": 5.615231954762322e-06, + "loss": 0.0124, + "step": 87530 + }, + { + "epoch": 25.469886528949665, + "grad_norm": 0.12661266326904297, + "learning_rate": 5.612007619559594e-06, + "loss": 0.0101, + "step": 87540 + }, + { + "epoch": 25.472796043060807, + "grad_norm": 0.09870382398366928, + "learning_rate": 5.608783997312932e-06, + "loss": 0.0123, + "step": 87550 + }, + { + "epoch": 25.475705557171953, + "grad_norm": 0.09002383798360825, + "learning_rate": 5.605561088267154e-06, + "loss": 0.0116, + "step": 87560 + }, + { + "epoch": 25.478615071283095, + "grad_norm": 0.10871091485023499, + "learning_rate": 5.602338892667007e-06, + "loss": 0.0113, + "step": 87570 + }, + { + "epoch": 25.48152458539424, + "grad_norm": 0.11083795130252838, + "learning_rate": 5.599117410757203e-06, + "loss": 0.0113, + "step": 87580 + }, + { + "epoch": 25.484434099505382, + "grad_norm": 0.07435361295938492, + "learning_rate": 5.595896642782397e-06, + "loss": 0.0123, + "step": 87590 + }, + { + "epoch": 25.487343613616527, + "grad_norm": 0.15081092715263367, + "learning_rate": 5.592676588987174e-06, + "loss": 0.0137, + "step": 87600 + }, + { + "epoch": 25.49025312772767, + "grad_norm": 0.17939072847366333, + "learning_rate": 5.589457249616086e-06, + "loss": 0.0113, + "step": 87610 + }, + { + "epoch": 25.49316264183881, + "grad_norm": 0.1547721028327942, + "learning_rate": 5.58623862491361e-06, + "loss": 0.0138, + "step": 87620 + }, + { + "epoch": 25.496072155949957, + "grad_norm": 0.1277560591697693, + "learning_rate": 5.583020715124191e-06, + "loss": 0.0126, + "step": 87630 + }, + { + "epoch": 25.4989816700611, + "grad_norm": 0.13131898641586304, + "learning_rate": 5.579803520492197e-06, + "loss": 0.0127, + "step": 87640 + }, + { + "epoch": 25.501891184172244, + "grad_norm": 0.10803813487291336, + "learning_rate": 5.576587041261962e-06, + "loss": 0.0122, + "step": 87650 + }, + { + "epoch": 25.504800698283386, + "grad_norm": 0.09478063881397247, + "learning_rate": 5.573371277677748e-06, + "loss": 0.0135, + "step": 87660 + }, + { + "epoch": 25.50771021239453, + "grad_norm": 0.08231937140226364, + "learning_rate": 5.5701562299837745e-06, + "loss": 0.012, + "step": 87670 + }, + { + "epoch": 25.510619726505674, + "grad_norm": 0.11144640296697617, + "learning_rate": 5.566941898424208e-06, + "loss": 0.011, + "step": 87680 + }, + { + "epoch": 25.513529240616815, + "grad_norm": 0.11873196810483932, + "learning_rate": 5.5637282832431476e-06, + "loss": 0.0112, + "step": 87690 + }, + { + "epoch": 25.51643875472796, + "grad_norm": 0.07882092893123627, + "learning_rate": 5.56051538468465e-06, + "loss": 0.0115, + "step": 87700 + }, + { + "epoch": 25.519348268839103, + "grad_norm": 0.10871482640504837, + "learning_rate": 5.557303202992719e-06, + "loss": 0.0112, + "step": 87710 + }, + { + "epoch": 25.52225778295025, + "grad_norm": 0.10573923587799072, + "learning_rate": 5.554091738411292e-06, + "loss": 0.011, + "step": 87720 + }, + { + "epoch": 25.52516729706139, + "grad_norm": 0.07802636921405792, + "learning_rate": 5.550880991184256e-06, + "loss": 0.0117, + "step": 87730 + }, + { + "epoch": 25.528076811172536, + "grad_norm": 0.11789896339178085, + "learning_rate": 5.5476709615554514e-06, + "loss": 0.0122, + "step": 87740 + }, + { + "epoch": 25.530986325283678, + "grad_norm": 0.11713097244501114, + "learning_rate": 5.5444616497686605e-06, + "loss": 0.0098, + "step": 87750 + }, + { + "epoch": 25.53389583939482, + "grad_norm": 0.09881651401519775, + "learning_rate": 5.5412530560676065e-06, + "loss": 0.0124, + "step": 87760 + }, + { + "epoch": 25.536805353505965, + "grad_norm": 0.10137613862752914, + "learning_rate": 5.5380451806959495e-06, + "loss": 0.0127, + "step": 87770 + }, + { + "epoch": 25.539714867617107, + "grad_norm": 0.11559659987688065, + "learning_rate": 5.534838023897329e-06, + "loss": 0.0116, + "step": 87780 + }, + { + "epoch": 25.542624381728253, + "grad_norm": 0.08491960912942886, + "learning_rate": 5.531631585915296e-06, + "loss": 0.0101, + "step": 87790 + }, + { + "epoch": 25.545533895839394, + "grad_norm": 0.11502115428447723, + "learning_rate": 5.528425866993353e-06, + "loss": 0.0122, + "step": 87800 + }, + { + "epoch": 25.54844340995054, + "grad_norm": 0.06962329149246216, + "learning_rate": 5.525220867374957e-06, + "loss": 0.0117, + "step": 87810 + }, + { + "epoch": 25.551352924061682, + "grad_norm": 0.1086135283112526, + "learning_rate": 5.5220165873035156e-06, + "loss": 0.0119, + "step": 87820 + }, + { + "epoch": 25.554262438172824, + "grad_norm": 0.10104059427976608, + "learning_rate": 5.518813027022363e-06, + "loss": 0.0131, + "step": 87830 + }, + { + "epoch": 25.55717195228397, + "grad_norm": 0.1369713544845581, + "learning_rate": 5.515610186774787e-06, + "loss": 0.0101, + "step": 87840 + }, + { + "epoch": 25.56008146639511, + "grad_norm": 0.11611921340227127, + "learning_rate": 5.512408066804025e-06, + "loss": 0.0125, + "step": 87850 + }, + { + "epoch": 25.562990980506257, + "grad_norm": 0.08678068965673447, + "learning_rate": 5.509206667353264e-06, + "loss": 0.0108, + "step": 87860 + }, + { + "epoch": 25.5659004946174, + "grad_norm": 0.09598438441753387, + "learning_rate": 5.506005988665617e-06, + "loss": 0.0126, + "step": 87870 + }, + { + "epoch": 25.56881000872854, + "grad_norm": 0.09993021935224533, + "learning_rate": 5.502806030984159e-06, + "loss": 0.011, + "step": 87880 + }, + { + "epoch": 25.571719522839686, + "grad_norm": 0.12174860388040543, + "learning_rate": 5.499606794551913e-06, + "loss": 0.0112, + "step": 87890 + }, + { + "epoch": 25.574629036950828, + "grad_norm": 0.11917684227228165, + "learning_rate": 5.496408279611833e-06, + "loss": 0.0118, + "step": 87900 + }, + { + "epoch": 25.577538551061973, + "grad_norm": 0.11314687132835388, + "learning_rate": 5.493210486406823e-06, + "loss": 0.0101, + "step": 87910 + }, + { + "epoch": 25.580448065173115, + "grad_norm": 0.12913261353969574, + "learning_rate": 5.490013415179735e-06, + "loss": 0.0116, + "step": 87920 + }, + { + "epoch": 25.58335757928426, + "grad_norm": 0.11864857375621796, + "learning_rate": 5.486817066173374e-06, + "loss": 0.0104, + "step": 87930 + }, + { + "epoch": 25.586267093395403, + "grad_norm": 0.10822846740484238, + "learning_rate": 5.483621439630469e-06, + "loss": 0.0124, + "step": 87940 + }, + { + "epoch": 25.589176607506545, + "grad_norm": 0.11641058325767517, + "learning_rate": 5.480426535793719e-06, + "loss": 0.0121, + "step": 87950 + }, + { + "epoch": 25.59208612161769, + "grad_norm": 0.10689792037010193, + "learning_rate": 5.477232354905746e-06, + "loss": 0.011, + "step": 87960 + }, + { + "epoch": 25.594995635728832, + "grad_norm": 0.11325965821743011, + "learning_rate": 5.474038897209129e-06, + "loss": 0.0109, + "step": 87970 + }, + { + "epoch": 25.597905149839978, + "grad_norm": 0.09292846918106079, + "learning_rate": 5.470846162946398e-06, + "loss": 0.0118, + "step": 87980 + }, + { + "epoch": 25.60081466395112, + "grad_norm": 0.08531259000301361, + "learning_rate": 5.4676541523600075e-06, + "loss": 0.0101, + "step": 87990 + }, + { + "epoch": 25.603724178062265, + "grad_norm": 0.08686395734548569, + "learning_rate": 5.464462865692382e-06, + "loss": 0.0113, + "step": 88000 + }, + { + "epoch": 25.606633692173407, + "grad_norm": 0.11074709892272949, + "learning_rate": 5.461272303185868e-06, + "loss": 0.0131, + "step": 88010 + }, + { + "epoch": 25.60954320628455, + "grad_norm": 0.11511655896902084, + "learning_rate": 5.458082465082778e-06, + "loss": 0.0123, + "step": 88020 + }, + { + "epoch": 25.612452720395694, + "grad_norm": 0.10177662968635559, + "learning_rate": 5.454893351625348e-06, + "loss": 0.0115, + "step": 88030 + }, + { + "epoch": 25.615362234506836, + "grad_norm": 0.06893067806959152, + "learning_rate": 5.4517049630557756e-06, + "loss": 0.011, + "step": 88040 + }, + { + "epoch": 25.618271748617982, + "grad_norm": 0.0794714167714119, + "learning_rate": 5.448517299616203e-06, + "loss": 0.0107, + "step": 88050 + }, + { + "epoch": 25.621181262729124, + "grad_norm": 0.08444253355264664, + "learning_rate": 5.445330361548707e-06, + "loss": 0.0128, + "step": 88060 + }, + { + "epoch": 25.62409077684027, + "grad_norm": 0.1005539670586586, + "learning_rate": 5.442144149095304e-06, + "loss": 0.0122, + "step": 88070 + }, + { + "epoch": 25.62700029095141, + "grad_norm": 0.10637499392032623, + "learning_rate": 5.438958662497988e-06, + "loss": 0.015, + "step": 88080 + }, + { + "epoch": 25.629909805062553, + "grad_norm": 0.11277727037668228, + "learning_rate": 5.435773901998663e-06, + "loss": 0.0117, + "step": 88090 + }, + { + "epoch": 25.6328193191737, + "grad_norm": 0.08572866022586823, + "learning_rate": 5.432589867839189e-06, + "loss": 0.0131, + "step": 88100 + }, + { + "epoch": 25.63572883328484, + "grad_norm": 0.15232455730438232, + "learning_rate": 5.429406560261372e-06, + "loss": 0.0142, + "step": 88110 + }, + { + "epoch": 25.638638347395986, + "grad_norm": 0.0767337903380394, + "learning_rate": 5.4262239795069734e-06, + "loss": 0.0114, + "step": 88120 + }, + { + "epoch": 25.641547861507128, + "grad_norm": 0.12770430743694305, + "learning_rate": 5.423042125817681e-06, + "loss": 0.012, + "step": 88130 + }, + { + "epoch": 25.644457375618273, + "grad_norm": 0.09571942687034607, + "learning_rate": 5.419860999435133e-06, + "loss": 0.0109, + "step": 88140 + }, + { + "epoch": 25.647366889729415, + "grad_norm": 0.09836230427026749, + "learning_rate": 5.4166806006009175e-06, + "loss": 0.0115, + "step": 88150 + }, + { + "epoch": 25.650276403840557, + "grad_norm": 0.103660948574543, + "learning_rate": 5.413500929556572e-06, + "loss": 0.0104, + "step": 88160 + }, + { + "epoch": 25.653185917951703, + "grad_norm": 0.08697967976331711, + "learning_rate": 5.410321986543565e-06, + "loss": 0.0128, + "step": 88170 + }, + { + "epoch": 25.656095432062845, + "grad_norm": 0.14189639687538147, + "learning_rate": 5.407143771803305e-06, + "loss": 0.0095, + "step": 88180 + }, + { + "epoch": 25.65900494617399, + "grad_norm": 0.18440014123916626, + "learning_rate": 5.403966285577181e-06, + "loss": 0.0109, + "step": 88190 + }, + { + "epoch": 25.661914460285132, + "grad_norm": 0.1162438839673996, + "learning_rate": 5.4007895281064874e-06, + "loss": 0.0117, + "step": 88200 + }, + { + "epoch": 25.664823974396278, + "grad_norm": 0.09603282809257507, + "learning_rate": 5.397613499632476e-06, + "loss": 0.0124, + "step": 88210 + }, + { + "epoch": 25.66773348850742, + "grad_norm": 0.11100269109010696, + "learning_rate": 5.39443820039635e-06, + "loss": 0.0115, + "step": 88220 + }, + { + "epoch": 25.67064300261856, + "grad_norm": 0.11508617550134659, + "learning_rate": 5.391263630639255e-06, + "loss": 0.0145, + "step": 88230 + }, + { + "epoch": 25.673552516729707, + "grad_norm": 0.08719713240861893, + "learning_rate": 5.38808979060227e-06, + "loss": 0.0108, + "step": 88240 + }, + { + "epoch": 25.67646203084085, + "grad_norm": 0.12943428754806519, + "learning_rate": 5.384916680526438e-06, + "loss": 0.0127, + "step": 88250 + }, + { + "epoch": 25.679371544951994, + "grad_norm": 0.08845295757055283, + "learning_rate": 5.381744300652724e-06, + "loss": 0.0101, + "step": 88260 + }, + { + "epoch": 25.682281059063136, + "grad_norm": 0.11877479404211044, + "learning_rate": 5.37857265122206e-06, + "loss": 0.0116, + "step": 88270 + }, + { + "epoch": 25.68519057317428, + "grad_norm": 0.08659867942333221, + "learning_rate": 5.375401732475303e-06, + "loss": 0.0113, + "step": 88280 + }, + { + "epoch": 25.688100087285424, + "grad_norm": 0.08103295415639877, + "learning_rate": 5.372231544653269e-06, + "loss": 0.0132, + "step": 88290 + }, + { + "epoch": 25.691009601396566, + "grad_norm": 0.11643160134553909, + "learning_rate": 5.369062087996716e-06, + "loss": 0.014, + "step": 88300 + }, + { + "epoch": 25.69391911550771, + "grad_norm": 0.11680260300636292, + "learning_rate": 5.3658933627463334e-06, + "loss": 0.0111, + "step": 88310 + }, + { + "epoch": 25.696828629618853, + "grad_norm": 0.11752639710903168, + "learning_rate": 5.362725369142775e-06, + "loss": 0.0128, + "step": 88320 + }, + { + "epoch": 25.69973814373, + "grad_norm": 0.10498029738664627, + "learning_rate": 5.359558107426618e-06, + "loss": 0.0099, + "step": 88330 + }, + { + "epoch": 25.70264765784114, + "grad_norm": 0.07952028512954712, + "learning_rate": 5.3563915778384046e-06, + "loss": 0.0123, + "step": 88340 + }, + { + "epoch": 25.705557171952282, + "grad_norm": 0.13650281727313995, + "learning_rate": 5.353225780618612e-06, + "loss": 0.0141, + "step": 88350 + }, + { + "epoch": 25.708466686063428, + "grad_norm": 0.10780907422304153, + "learning_rate": 5.350060716007657e-06, + "loss": 0.0111, + "step": 88360 + }, + { + "epoch": 25.71137620017457, + "grad_norm": 0.0968942940235138, + "learning_rate": 5.346896384245902e-06, + "loss": 0.0115, + "step": 88370 + }, + { + "epoch": 25.714285714285715, + "grad_norm": 0.15827828645706177, + "learning_rate": 5.3437327855736615e-06, + "loss": 0.0109, + "step": 88380 + }, + { + "epoch": 25.717195228396857, + "grad_norm": 0.10371927917003632, + "learning_rate": 5.340569920231196e-06, + "loss": 0.0123, + "step": 88390 + }, + { + "epoch": 25.720104742508003, + "grad_norm": 0.1169683113694191, + "learning_rate": 5.337407788458694e-06, + "loss": 0.0139, + "step": 88400 + }, + { + "epoch": 25.723014256619145, + "grad_norm": 0.10406654328107834, + "learning_rate": 5.334246390496302e-06, + "loss": 0.0109, + "step": 88410 + }, + { + "epoch": 25.725923770730287, + "grad_norm": 0.0923481360077858, + "learning_rate": 5.331085726584114e-06, + "loss": 0.0115, + "step": 88420 + }, + { + "epoch": 25.728833284841432, + "grad_norm": 0.14993013441562653, + "learning_rate": 5.327925796962155e-06, + "loss": 0.0123, + "step": 88430 + }, + { + "epoch": 25.731742798952574, + "grad_norm": 0.13100889325141907, + "learning_rate": 5.324766601870396e-06, + "loss": 0.0129, + "step": 88440 + }, + { + "epoch": 25.73465231306372, + "grad_norm": 0.09301172941923141, + "learning_rate": 5.321608141548763e-06, + "loss": 0.012, + "step": 88450 + }, + { + "epoch": 25.73756182717486, + "grad_norm": 0.09636744856834412, + "learning_rate": 5.318450416237125e-06, + "loss": 0.0116, + "step": 88460 + }, + { + "epoch": 25.740471341286007, + "grad_norm": 0.07884202152490616, + "learning_rate": 5.315293426175287e-06, + "loss": 0.0117, + "step": 88470 + }, + { + "epoch": 25.74338085539715, + "grad_norm": 0.08410536497831345, + "learning_rate": 5.312137171602987e-06, + "loss": 0.0118, + "step": 88480 + }, + { + "epoch": 25.74629036950829, + "grad_norm": 0.08902568370103836, + "learning_rate": 5.308981652759946e-06, + "loss": 0.0132, + "step": 88490 + }, + { + "epoch": 25.749199883619436, + "grad_norm": 0.14117811620235443, + "learning_rate": 5.305826869885792e-06, + "loss": 0.0102, + "step": 88500 + }, + { + "epoch": 25.752109397730578, + "grad_norm": 0.09212297946214676, + "learning_rate": 5.302672823220107e-06, + "loss": 0.0112, + "step": 88510 + }, + { + "epoch": 25.755018911841724, + "grad_norm": 0.10552917420864105, + "learning_rate": 5.299519513002425e-06, + "loss": 0.0118, + "step": 88520 + }, + { + "epoch": 25.757928425952866, + "grad_norm": 0.0877046287059784, + "learning_rate": 5.296366939472222e-06, + "loss": 0.0121, + "step": 88530 + }, + { + "epoch": 25.760837940064008, + "grad_norm": 0.11854823678731918, + "learning_rate": 5.293215102868914e-06, + "loss": 0.0134, + "step": 88540 + }, + { + "epoch": 25.763747454175153, + "grad_norm": 0.0997321754693985, + "learning_rate": 5.290064003431852e-06, + "loss": 0.0121, + "step": 88550 + }, + { + "epoch": 25.766656968286295, + "grad_norm": 0.10242005437612534, + "learning_rate": 5.286913641400351e-06, + "loss": 0.0115, + "step": 88560 + }, + { + "epoch": 25.76956648239744, + "grad_norm": 0.07567281275987625, + "learning_rate": 5.2837640170136605e-06, + "loss": 0.0117, + "step": 88570 + }, + { + "epoch": 25.772475996508582, + "grad_norm": 0.10252887010574341, + "learning_rate": 5.28061513051097e-06, + "loss": 0.0125, + "step": 88580 + }, + { + "epoch": 25.775385510619728, + "grad_norm": 0.07606921344995499, + "learning_rate": 5.277466982131414e-06, + "loss": 0.0102, + "step": 88590 + }, + { + "epoch": 25.77829502473087, + "grad_norm": 0.22677850723266602, + "learning_rate": 5.274319572114085e-06, + "loss": 0.0131, + "step": 88600 + }, + { + "epoch": 25.78120453884201, + "grad_norm": 0.10978946834802628, + "learning_rate": 5.271172900697997e-06, + "loss": 0.0117, + "step": 88610 + }, + { + "epoch": 25.784114052953157, + "grad_norm": 0.08862143009901047, + "learning_rate": 5.268026968122125e-06, + "loss": 0.0146, + "step": 88620 + }, + { + "epoch": 25.7870235670643, + "grad_norm": 0.10283005982637405, + "learning_rate": 5.264881774625375e-06, + "loss": 0.0115, + "step": 88630 + }, + { + "epoch": 25.789933081175445, + "grad_norm": 0.17887049913406372, + "learning_rate": 5.261737320446613e-06, + "loss": 0.0113, + "step": 88640 + }, + { + "epoch": 25.792842595286587, + "grad_norm": 0.07957135885953903, + "learning_rate": 5.2585936058246284e-06, + "loss": 0.0105, + "step": 88650 + }, + { + "epoch": 25.795752109397732, + "grad_norm": 0.10643641650676727, + "learning_rate": 5.255450630998179e-06, + "loss": 0.0109, + "step": 88660 + }, + { + "epoch": 25.798661623508874, + "grad_norm": 0.08920148015022278, + "learning_rate": 5.252308396205938e-06, + "loss": 0.0114, + "step": 88670 + }, + { + "epoch": 25.801571137620016, + "grad_norm": 0.09693536162376404, + "learning_rate": 5.249166901686548e-06, + "loss": 0.0111, + "step": 88680 + }, + { + "epoch": 25.80448065173116, + "grad_norm": 0.1283738762140274, + "learning_rate": 5.2460261476785855e-06, + "loss": 0.0123, + "step": 88690 + }, + { + "epoch": 25.807390165842303, + "grad_norm": 0.08173682540655136, + "learning_rate": 5.242886134420562e-06, + "loss": 0.0118, + "step": 88700 + }, + { + "epoch": 25.81029967995345, + "grad_norm": 0.10182774811983109, + "learning_rate": 5.2397468621509455e-06, + "loss": 0.0114, + "step": 88710 + }, + { + "epoch": 25.81320919406459, + "grad_norm": 0.10274915397167206, + "learning_rate": 5.236608331108147e-06, + "loss": 0.0128, + "step": 88720 + }, + { + "epoch": 25.816118708175736, + "grad_norm": 0.09294171631336212, + "learning_rate": 5.2334705415305125e-06, + "loss": 0.0131, + "step": 88730 + }, + { + "epoch": 25.819028222286878, + "grad_norm": 0.11129912734031677, + "learning_rate": 5.230333493656333e-06, + "loss": 0.0131, + "step": 88740 + }, + { + "epoch": 25.82193773639802, + "grad_norm": 0.12304433435201645, + "learning_rate": 5.22719718772385e-06, + "loss": 0.0127, + "step": 88750 + }, + { + "epoch": 25.824847250509166, + "grad_norm": 0.09368003904819489, + "learning_rate": 5.22406162397125e-06, + "loss": 0.013, + "step": 88760 + }, + { + "epoch": 25.827756764620307, + "grad_norm": 0.2238706648349762, + "learning_rate": 5.220926802636656e-06, + "loss": 0.0119, + "step": 88770 + }, + { + "epoch": 25.830666278731453, + "grad_norm": 0.11115951091051102, + "learning_rate": 5.217792723958123e-06, + "loss": 0.0121, + "step": 88780 + }, + { + "epoch": 25.833575792842595, + "grad_norm": 0.10679587721824646, + "learning_rate": 5.2146593881736865e-06, + "loss": 0.0104, + "step": 88790 + }, + { + "epoch": 25.83648530695374, + "grad_norm": 0.10187658667564392, + "learning_rate": 5.211526795521293e-06, + "loss": 0.0123, + "step": 88800 + }, + { + "epoch": 25.839394821064882, + "grad_norm": 0.10133921355009079, + "learning_rate": 5.2083949462388346e-06, + "loss": 0.0107, + "step": 88810 + }, + { + "epoch": 25.842304335176024, + "grad_norm": 0.12432372570037842, + "learning_rate": 5.205263840564161e-06, + "loss": 0.011, + "step": 88820 + }, + { + "epoch": 25.84521384928717, + "grad_norm": 0.11117817461490631, + "learning_rate": 5.202133478735065e-06, + "loss": 0.009, + "step": 88830 + }, + { + "epoch": 25.84812336339831, + "grad_norm": 0.09145641326904297, + "learning_rate": 5.199003860989271e-06, + "loss": 0.0133, + "step": 88840 + }, + { + "epoch": 25.851032877509457, + "grad_norm": 0.23598967492580414, + "learning_rate": 5.1958749875644455e-06, + "loss": 0.0133, + "step": 88850 + }, + { + "epoch": 25.8539423916206, + "grad_norm": 0.10682088136672974, + "learning_rate": 5.192746858698216e-06, + "loss": 0.0129, + "step": 88860 + }, + { + "epoch": 25.856851905731745, + "grad_norm": 0.09650710970163345, + "learning_rate": 5.189619474628144e-06, + "loss": 0.0136, + "step": 88870 + }, + { + "epoch": 25.859761419842886, + "grad_norm": 0.10322786122560501, + "learning_rate": 5.186492835591725e-06, + "loss": 0.0136, + "step": 88880 + }, + { + "epoch": 25.86267093395403, + "grad_norm": 0.114606574177742, + "learning_rate": 5.183366941826413e-06, + "loss": 0.0138, + "step": 88890 + }, + { + "epoch": 25.865580448065174, + "grad_norm": 0.12629081308841705, + "learning_rate": 5.1802417935696e-06, + "loss": 0.0103, + "step": 88900 + }, + { + "epoch": 25.868489962176316, + "grad_norm": 0.07543253898620605, + "learning_rate": 5.177117391058615e-06, + "loss": 0.011, + "step": 88910 + }, + { + "epoch": 25.87139947628746, + "grad_norm": 0.10297894477844238, + "learning_rate": 5.1739937345307435e-06, + "loss": 0.0132, + "step": 88920 + }, + { + "epoch": 25.874308990398603, + "grad_norm": 0.11241215467453003, + "learning_rate": 5.1708708242231965e-06, + "loss": 0.0134, + "step": 88930 + }, + { + "epoch": 25.87721850450975, + "grad_norm": 0.14580558240413666, + "learning_rate": 5.16774866037315e-06, + "loss": 0.0124, + "step": 88940 + }, + { + "epoch": 25.88012801862089, + "grad_norm": 0.10049839317798615, + "learning_rate": 5.164627243217698e-06, + "loss": 0.0129, + "step": 88950 + }, + { + "epoch": 25.883037532732033, + "grad_norm": 0.14725244045257568, + "learning_rate": 5.161506572993907e-06, + "loss": 0.0115, + "step": 88960 + }, + { + "epoch": 25.885947046843178, + "grad_norm": 0.08832670748233795, + "learning_rate": 5.158386649938756e-06, + "loss": 0.0123, + "step": 88970 + }, + { + "epoch": 25.88885656095432, + "grad_norm": 0.08706319332122803, + "learning_rate": 5.1552674742891906e-06, + "loss": 0.0097, + "step": 88980 + }, + { + "epoch": 25.891766075065465, + "grad_norm": 0.08671584725379944, + "learning_rate": 5.152149046282097e-06, + "loss": 0.0095, + "step": 88990 + }, + { + "epoch": 25.894675589176607, + "grad_norm": 0.11073006689548492, + "learning_rate": 5.1490313661542886e-06, + "loss": 0.0114, + "step": 89000 + }, + { + "epoch": 25.89758510328775, + "grad_norm": 0.14182555675506592, + "learning_rate": 5.14591443414254e-06, + "loss": 0.0139, + "step": 89010 + }, + { + "epoch": 25.900494617398895, + "grad_norm": 0.17721961438655853, + "learning_rate": 5.142798250483555e-06, + "loss": 0.0117, + "step": 89020 + }, + { + "epoch": 25.903404131510037, + "grad_norm": 0.08609306067228317, + "learning_rate": 5.139682815413996e-06, + "loss": 0.0131, + "step": 89030 + }, + { + "epoch": 25.906313645621182, + "grad_norm": 0.09551149606704712, + "learning_rate": 5.136568129170449e-06, + "loss": 0.0119, + "step": 89040 + }, + { + "epoch": 25.909223159732324, + "grad_norm": 0.10553261637687683, + "learning_rate": 5.1334541919894594e-06, + "loss": 0.014, + "step": 89050 + }, + { + "epoch": 25.91213267384347, + "grad_norm": 0.12800158560276031, + "learning_rate": 5.130341004107518e-06, + "loss": 0.0113, + "step": 89060 + }, + { + "epoch": 25.91504218795461, + "grad_norm": 0.0991319790482521, + "learning_rate": 5.12722856576104e-06, + "loss": 0.0147, + "step": 89070 + }, + { + "epoch": 25.917951702065753, + "grad_norm": 0.09480152279138565, + "learning_rate": 5.1241168771863895e-06, + "loss": 0.0115, + "step": 89080 + }, + { + "epoch": 25.9208612161769, + "grad_norm": 0.11939186602830887, + "learning_rate": 5.121005938619897e-06, + "loss": 0.0142, + "step": 89090 + }, + { + "epoch": 25.92377073028804, + "grad_norm": 0.08479515463113785, + "learning_rate": 5.117895750297807e-06, + "loss": 0.0133, + "step": 89100 + }, + { + "epoch": 25.926680244399186, + "grad_norm": 0.1367853432893753, + "learning_rate": 5.1147863124563145e-06, + "loss": 0.0114, + "step": 89110 + }, + { + "epoch": 25.92958975851033, + "grad_norm": 0.16290047764778137, + "learning_rate": 5.111677625331565e-06, + "loss": 0.0132, + "step": 89120 + }, + { + "epoch": 25.932499272621474, + "grad_norm": 0.0941636934876442, + "learning_rate": 5.108569689159646e-06, + "loss": 0.0135, + "step": 89130 + }, + { + "epoch": 25.935408786732616, + "grad_norm": 0.05854180455207825, + "learning_rate": 5.105462504176583e-06, + "loss": 0.0106, + "step": 89140 + }, + { + "epoch": 25.938318300843758, + "grad_norm": 0.07304222136735916, + "learning_rate": 5.102356070618338e-06, + "loss": 0.0097, + "step": 89150 + }, + { + "epoch": 25.941227814954903, + "grad_norm": 0.1427576243877411, + "learning_rate": 5.099250388720832e-06, + "loss": 0.0099, + "step": 89160 + }, + { + "epoch": 25.944137329066045, + "grad_norm": 0.10147321224212646, + "learning_rate": 5.096145458719924e-06, + "loss": 0.014, + "step": 89170 + }, + { + "epoch": 25.94704684317719, + "grad_norm": 0.09261626750230789, + "learning_rate": 5.0930412808514045e-06, + "loss": 0.0119, + "step": 89180 + }, + { + "epoch": 25.949956357288332, + "grad_norm": 0.07010843604803085, + "learning_rate": 5.089937855351019e-06, + "loss": 0.0124, + "step": 89190 + }, + { + "epoch": 25.952865871399478, + "grad_norm": 0.07714640349149704, + "learning_rate": 5.086835182454458e-06, + "loss": 0.011, + "step": 89200 + }, + { + "epoch": 25.95577538551062, + "grad_norm": 0.10557691007852554, + "learning_rate": 5.083733262397342e-06, + "loss": 0.0113, + "step": 89210 + }, + { + "epoch": 25.958684899621762, + "grad_norm": 0.10042458027601242, + "learning_rate": 5.0806320954152394e-06, + "loss": 0.0122, + "step": 89220 + }, + { + "epoch": 25.961594413732907, + "grad_norm": 0.12709765136241913, + "learning_rate": 5.077531681743667e-06, + "loss": 0.0122, + "step": 89230 + }, + { + "epoch": 25.96450392784405, + "grad_norm": 0.12762664258480072, + "learning_rate": 5.074432021618086e-06, + "loss": 0.0124, + "step": 89240 + }, + { + "epoch": 25.967413441955195, + "grad_norm": 0.0882437527179718, + "learning_rate": 5.071333115273885e-06, + "loss": 0.0123, + "step": 89250 + }, + { + "epoch": 25.970322956066337, + "grad_norm": 0.13078820705413818, + "learning_rate": 5.068234962946416e-06, + "loss": 0.0117, + "step": 89260 + }, + { + "epoch": 25.97323247017748, + "grad_norm": 0.07908745855093002, + "learning_rate": 5.065137564870952e-06, + "loss": 0.0117, + "step": 89270 + }, + { + "epoch": 25.976141984288624, + "grad_norm": 0.063503198325634, + "learning_rate": 5.062040921282725e-06, + "loss": 0.0103, + "step": 89280 + }, + { + "epoch": 25.979051498399766, + "grad_norm": 0.10239052772521973, + "learning_rate": 5.058945032416909e-06, + "loss": 0.0115, + "step": 89290 + }, + { + "epoch": 25.98196101251091, + "grad_norm": 0.10790276527404785, + "learning_rate": 5.05584989850861e-06, + "loss": 0.0107, + "step": 89300 + }, + { + "epoch": 25.984870526622053, + "grad_norm": 0.1208251416683197, + "learning_rate": 5.052755519792889e-06, + "loss": 0.0092, + "step": 89310 + }, + { + "epoch": 25.9877800407332, + "grad_norm": 0.10080737620592117, + "learning_rate": 5.049661896504736e-06, + "loss": 0.012, + "step": 89320 + }, + { + "epoch": 25.99068955484434, + "grad_norm": 0.10869285464286804, + "learning_rate": 5.0465690288791e-06, + "loss": 0.0102, + "step": 89330 + }, + { + "epoch": 25.993599068955483, + "grad_norm": 0.11024977266788483, + "learning_rate": 5.043476917150854e-06, + "loss": 0.0103, + "step": 89340 + }, + { + "epoch": 25.99650858306663, + "grad_norm": 0.07004570215940475, + "learning_rate": 5.040385561554829e-06, + "loss": 0.0104, + "step": 89350 + }, + { + "epoch": 25.99941809717777, + "grad_norm": 0.11394871771335602, + "learning_rate": 5.037294962325799e-06, + "loss": 0.0102, + "step": 89360 + }, + { + "epoch": 26.002327611288916, + "grad_norm": 0.06968764215707779, + "learning_rate": 5.034205119698466e-06, + "loss": 0.0112, + "step": 89370 + }, + { + "epoch": 26.005237125400058, + "grad_norm": 0.0978279635310173, + "learning_rate": 5.031116033907483e-06, + "loss": 0.0118, + "step": 89380 + }, + { + "epoch": 26.008146639511203, + "grad_norm": 0.08756644278764725, + "learning_rate": 5.028027705187448e-06, + "loss": 0.0112, + "step": 89390 + }, + { + "epoch": 26.011056153622345, + "grad_norm": 0.10527973622083664, + "learning_rate": 5.024940133772904e-06, + "loss": 0.0122, + "step": 89400 + }, + { + "epoch": 26.013965667733487, + "grad_norm": 0.14659753441810608, + "learning_rate": 5.0218533198983215e-06, + "loss": 0.0126, + "step": 89410 + }, + { + "epoch": 26.016875181844632, + "grad_norm": 0.12372612208127975, + "learning_rate": 5.018767263798132e-06, + "loss": 0.0124, + "step": 89420 + }, + { + "epoch": 26.019784695955774, + "grad_norm": 0.09608685970306396, + "learning_rate": 5.0156819657067e-06, + "loss": 0.0119, + "step": 89430 + }, + { + "epoch": 26.02269421006692, + "grad_norm": 0.12133630365133286, + "learning_rate": 5.012597425858334e-06, + "loss": 0.0133, + "step": 89440 + }, + { + "epoch": 26.025603724178062, + "grad_norm": 0.09444485604763031, + "learning_rate": 5.009513644487276e-06, + "loss": 0.011, + "step": 89450 + }, + { + "epoch": 26.028513238289207, + "grad_norm": 0.10371816158294678, + "learning_rate": 5.006430621827726e-06, + "loss": 0.0108, + "step": 89460 + }, + { + "epoch": 26.03142275240035, + "grad_norm": 0.07457626610994339, + "learning_rate": 5.003348358113824e-06, + "loss": 0.0097, + "step": 89470 + }, + { + "epoch": 26.03433226651149, + "grad_norm": 0.12135842442512512, + "learning_rate": 5.000266853579641e-06, + "loss": 0.0107, + "step": 89480 + }, + { + "epoch": 26.037241780622637, + "grad_norm": 0.136338010430336, + "learning_rate": 4.997186108459188e-06, + "loss": 0.0125, + "step": 89490 + }, + { + "epoch": 26.04015129473378, + "grad_norm": 0.07690443098545074, + "learning_rate": 4.994106122986448e-06, + "loss": 0.011, + "step": 89500 + }, + { + "epoch": 26.043060808844924, + "grad_norm": 0.08546865731477737, + "learning_rate": 4.991026897395314e-06, + "loss": 0.0097, + "step": 89510 + }, + { + "epoch": 26.045970322956066, + "grad_norm": 0.10333948582410812, + "learning_rate": 4.98794843191963e-06, + "loss": 0.01, + "step": 89520 + }, + { + "epoch": 26.04887983706721, + "grad_norm": 0.0968593955039978, + "learning_rate": 4.984870726793188e-06, + "loss": 0.0087, + "step": 89530 + }, + { + "epoch": 26.051789351178353, + "grad_norm": 0.09621843695640564, + "learning_rate": 4.981793782249725e-06, + "loss": 0.0156, + "step": 89540 + }, + { + "epoch": 26.054698865289495, + "grad_norm": 0.10237745940685272, + "learning_rate": 4.978717598522905e-06, + "loss": 0.011, + "step": 89550 + }, + { + "epoch": 26.05760837940064, + "grad_norm": 0.08376898616552353, + "learning_rate": 4.975642175846352e-06, + "loss": 0.0105, + "step": 89560 + }, + { + "epoch": 26.060517893511783, + "grad_norm": 0.08209507167339325, + "learning_rate": 4.97256751445362e-06, + "loss": 0.0115, + "step": 89570 + }, + { + "epoch": 26.063427407622928, + "grad_norm": 0.10036876052618027, + "learning_rate": 4.969493614578211e-06, + "loss": 0.0134, + "step": 89580 + }, + { + "epoch": 26.06633692173407, + "grad_norm": 0.12767066061496735, + "learning_rate": 4.966420476453563e-06, + "loss": 0.0103, + "step": 89590 + }, + { + "epoch": 26.069246435845216, + "grad_norm": 0.12127848714590073, + "learning_rate": 4.963348100313064e-06, + "loss": 0.013, + "step": 89600 + }, + { + "epoch": 26.072155949956358, + "grad_norm": 0.1078428328037262, + "learning_rate": 4.9602764863900445e-06, + "loss": 0.0114, + "step": 89610 + }, + { + "epoch": 26.0750654640675, + "grad_norm": 0.07218681275844574, + "learning_rate": 4.957205634917766e-06, + "loss": 0.0115, + "step": 89620 + }, + { + "epoch": 26.077974978178645, + "grad_norm": 0.08556217700242996, + "learning_rate": 4.954135546129448e-06, + "loss": 0.0124, + "step": 89630 + }, + { + "epoch": 26.080884492289787, + "grad_norm": 0.09663402289152145, + "learning_rate": 4.951066220258231e-06, + "loss": 0.011, + "step": 89640 + }, + { + "epoch": 26.083794006400932, + "grad_norm": 0.1496475636959076, + "learning_rate": 4.947997657537219e-06, + "loss": 0.0107, + "step": 89650 + }, + { + "epoch": 26.086703520512074, + "grad_norm": 0.08986788243055344, + "learning_rate": 4.944929858199452e-06, + "loss": 0.0116, + "step": 89660 + }, + { + "epoch": 26.089613034623216, + "grad_norm": 0.0975010022521019, + "learning_rate": 4.941862822477905e-06, + "loss": 0.0129, + "step": 89670 + }, + { + "epoch": 26.09252254873436, + "grad_norm": 0.1189657673239708, + "learning_rate": 4.938796550605494e-06, + "loss": 0.0105, + "step": 89680 + }, + { + "epoch": 26.095432062845504, + "grad_norm": 0.14361122250556946, + "learning_rate": 4.935731042815088e-06, + "loss": 0.0094, + "step": 89690 + }, + { + "epoch": 26.09834157695665, + "grad_norm": 0.11967582255601883, + "learning_rate": 4.932666299339494e-06, + "loss": 0.0107, + "step": 89700 + }, + { + "epoch": 26.10125109106779, + "grad_norm": 0.07836959511041641, + "learning_rate": 4.929602320411451e-06, + "loss": 0.011, + "step": 89710 + }, + { + "epoch": 26.104160605178937, + "grad_norm": 0.08054110407829285, + "learning_rate": 4.926539106263653e-06, + "loss": 0.0125, + "step": 89720 + }, + { + "epoch": 26.10707011929008, + "grad_norm": 0.17746509611606598, + "learning_rate": 4.923476657128736e-06, + "loss": 0.0117, + "step": 89730 + }, + { + "epoch": 26.10997963340122, + "grad_norm": 0.10785964876413345, + "learning_rate": 4.920414973239267e-06, + "loss": 0.0123, + "step": 89740 + }, + { + "epoch": 26.112889147512366, + "grad_norm": 0.14870581030845642, + "learning_rate": 4.917354054827757e-06, + "loss": 0.0106, + "step": 89750 + }, + { + "epoch": 26.115798661623508, + "grad_norm": 0.07345924526453018, + "learning_rate": 4.914293902126668e-06, + "loss": 0.0112, + "step": 89760 + }, + { + "epoch": 26.118708175734653, + "grad_norm": 0.08498101681470871, + "learning_rate": 4.9112345153684005e-06, + "loss": 0.0119, + "step": 89770 + }, + { + "epoch": 26.121617689845795, + "grad_norm": 0.08542249351739883, + "learning_rate": 4.908175894785293e-06, + "loss": 0.0134, + "step": 89780 + }, + { + "epoch": 26.12452720395694, + "grad_norm": 0.11035823076963425, + "learning_rate": 4.905118040609615e-06, + "loss": 0.0122, + "step": 89790 + }, + { + "epoch": 26.127436718068083, + "grad_norm": 0.10539872199296951, + "learning_rate": 4.9020609530736116e-06, + "loss": 0.011, + "step": 89800 + }, + { + "epoch": 26.130346232179225, + "grad_norm": 0.10010574758052826, + "learning_rate": 4.899004632409438e-06, + "loss": 0.0115, + "step": 89810 + }, + { + "epoch": 26.13325574629037, + "grad_norm": 0.0883035659790039, + "learning_rate": 4.895949078849196e-06, + "loss": 0.0108, + "step": 89820 + }, + { + "epoch": 26.136165260401512, + "grad_norm": 0.08121582120656967, + "learning_rate": 4.892894292624943e-06, + "loss": 0.0114, + "step": 89830 + }, + { + "epoch": 26.139074774512657, + "grad_norm": 0.12971028685569763, + "learning_rate": 4.8898402739686695e-06, + "loss": 0.0102, + "step": 89840 + }, + { + "epoch": 26.1419842886238, + "grad_norm": 0.1204080879688263, + "learning_rate": 4.886787023112308e-06, + "loss": 0.0104, + "step": 89850 + }, + { + "epoch": 26.144893802734945, + "grad_norm": 0.1499832719564438, + "learning_rate": 4.883734540287725e-06, + "loss": 0.0103, + "step": 89860 + }, + { + "epoch": 26.147803316846087, + "grad_norm": 0.12314704805612564, + "learning_rate": 4.880682825726742e-06, + "loss": 0.0133, + "step": 89870 + }, + { + "epoch": 26.15071283095723, + "grad_norm": 0.08866723626852036, + "learning_rate": 4.877631879661123e-06, + "loss": 0.0127, + "step": 89880 + }, + { + "epoch": 26.153622345068374, + "grad_norm": 0.17901040613651276, + "learning_rate": 4.874581702322553e-06, + "loss": 0.0101, + "step": 89890 + }, + { + "epoch": 26.156531859179516, + "grad_norm": 0.09928476065397263, + "learning_rate": 4.871532293942683e-06, + "loss": 0.011, + "step": 89900 + }, + { + "epoch": 26.15944137329066, + "grad_norm": 0.11077095568180084, + "learning_rate": 4.868483654753097e-06, + "loss": 0.011, + "step": 89910 + }, + { + "epoch": 26.162350887401804, + "grad_norm": 0.11131104826927185, + "learning_rate": 4.865435784985311e-06, + "loss": 0.0122, + "step": 89920 + }, + { + "epoch": 26.16526040151295, + "grad_norm": 0.10429958254098892, + "learning_rate": 4.8623886848708e-06, + "loss": 0.0131, + "step": 89930 + }, + { + "epoch": 26.16816991562409, + "grad_norm": 0.1763681173324585, + "learning_rate": 4.859342354640961e-06, + "loss": 0.0109, + "step": 89940 + }, + { + "epoch": 26.171079429735233, + "grad_norm": 0.09277047961950302, + "learning_rate": 4.856296794527153e-06, + "loss": 0.0116, + "step": 89950 + }, + { + "epoch": 26.17398894384638, + "grad_norm": 0.1342877596616745, + "learning_rate": 4.853252004760654e-06, + "loss": 0.0123, + "step": 89960 + }, + { + "epoch": 26.17689845795752, + "grad_norm": 0.09743522107601166, + "learning_rate": 4.85020798557271e-06, + "loss": 0.0116, + "step": 89970 + }, + { + "epoch": 26.179807972068666, + "grad_norm": 0.08068060129880905, + "learning_rate": 4.84716473719448e-06, + "loss": 0.0098, + "step": 89980 + }, + { + "epoch": 26.182717486179808, + "grad_norm": 0.1007978618144989, + "learning_rate": 4.844122259857086e-06, + "loss": 0.0107, + "step": 89990 + }, + { + "epoch": 26.18562700029095, + "grad_norm": 0.0839049369096756, + "learning_rate": 4.841080553791589e-06, + "loss": 0.0104, + "step": 90000 + }, + { + "epoch": 26.188536514402095, + "grad_norm": 0.08725588023662567, + "learning_rate": 4.838039619228974e-06, + "loss": 0.011, + "step": 90010 + }, + { + "epoch": 26.191446028513237, + "grad_norm": 0.1069418266415596, + "learning_rate": 4.834999456400188e-06, + "loss": 0.0109, + "step": 90020 + }, + { + "epoch": 26.194355542624383, + "grad_norm": 0.149014413356781, + "learning_rate": 4.831960065536115e-06, + "loss": 0.0113, + "step": 90030 + }, + { + "epoch": 26.197265056735525, + "grad_norm": 0.09270782023668289, + "learning_rate": 4.828921446867572e-06, + "loss": 0.011, + "step": 90040 + }, + { + "epoch": 26.20017457084667, + "grad_norm": 0.10007612407207489, + "learning_rate": 4.825883600625316e-06, + "loss": 0.0107, + "step": 90050 + }, + { + "epoch": 26.203084084957812, + "grad_norm": 0.11658672988414764, + "learning_rate": 4.822846527040058e-06, + "loss": 0.0115, + "step": 90060 + }, + { + "epoch": 26.205993599068954, + "grad_norm": 0.16269689798355103, + "learning_rate": 4.819810226342449e-06, + "loss": 0.0103, + "step": 90070 + }, + { + "epoch": 26.2089031131801, + "grad_norm": 0.11922622472047806, + "learning_rate": 4.816774698763067e-06, + "loss": 0.0109, + "step": 90080 + }, + { + "epoch": 26.21181262729124, + "grad_norm": 0.27552881836891174, + "learning_rate": 4.813739944532436e-06, + "loss": 0.0109, + "step": 90090 + }, + { + "epoch": 26.214722141402387, + "grad_norm": 0.08755198866128922, + "learning_rate": 4.810705963881042e-06, + "loss": 0.0118, + "step": 90100 + }, + { + "epoch": 26.21763165551353, + "grad_norm": 0.1502755731344223, + "learning_rate": 4.807672757039286e-06, + "loss": 0.0137, + "step": 90110 + }, + { + "epoch": 26.220541169624674, + "grad_norm": 0.19614984095096588, + "learning_rate": 4.8046403242375155e-06, + "loss": 0.0144, + "step": 90120 + }, + { + "epoch": 26.223450683735816, + "grad_norm": 0.11615870893001556, + "learning_rate": 4.801608665706029e-06, + "loss": 0.0105, + "step": 90130 + }, + { + "epoch": 26.226360197846958, + "grad_norm": 0.10994139313697815, + "learning_rate": 4.798577781675066e-06, + "loss": 0.0129, + "step": 90140 + }, + { + "epoch": 26.229269711958104, + "grad_norm": 0.10202840715646744, + "learning_rate": 4.795547672374798e-06, + "loss": 0.0116, + "step": 90150 + }, + { + "epoch": 26.232179226069245, + "grad_norm": 0.12712840735912323, + "learning_rate": 4.792518338035334e-06, + "loss": 0.015, + "step": 90160 + }, + { + "epoch": 26.23508874018039, + "grad_norm": 0.1922171711921692, + "learning_rate": 4.78948977888674e-06, + "loss": 0.0121, + "step": 90170 + }, + { + "epoch": 26.237998254291533, + "grad_norm": 0.13594913482666016, + "learning_rate": 4.786461995159017e-06, + "loss": 0.0099, + "step": 90180 + }, + { + "epoch": 26.24090776840268, + "grad_norm": 0.09519024938344955, + "learning_rate": 4.7834349870820984e-06, + "loss": 0.0114, + "step": 90190 + }, + { + "epoch": 26.24381728251382, + "grad_norm": 0.13181206583976746, + "learning_rate": 4.780408754885869e-06, + "loss": 0.011, + "step": 90200 + }, + { + "epoch": 26.246726796624962, + "grad_norm": 0.09033214300870895, + "learning_rate": 4.777383298800157e-06, + "loss": 0.0124, + "step": 90210 + }, + { + "epoch": 26.249636310736108, + "grad_norm": 0.1790747344493866, + "learning_rate": 4.774358619054718e-06, + "loss": 0.0121, + "step": 90220 + }, + { + "epoch": 26.25254582484725, + "grad_norm": 0.12913504242897034, + "learning_rate": 4.771334715879254e-06, + "loss": 0.0128, + "step": 90230 + }, + { + "epoch": 26.255455338958395, + "grad_norm": 0.10624168068170547, + "learning_rate": 4.768311589503417e-06, + "loss": 0.0122, + "step": 90240 + }, + { + "epoch": 26.258364853069537, + "grad_norm": 0.10724175721406937, + "learning_rate": 4.765289240156794e-06, + "loss": 0.0121, + "step": 90250 + }, + { + "epoch": 26.261274367180683, + "grad_norm": 0.08217649161815643, + "learning_rate": 4.762267668068906e-06, + "loss": 0.0119, + "step": 90260 + }, + { + "epoch": 26.264183881291824, + "grad_norm": 0.10530491173267365, + "learning_rate": 4.759246873469231e-06, + "loss": 0.0119, + "step": 90270 + }, + { + "epoch": 26.267093395402966, + "grad_norm": 0.09962466359138489, + "learning_rate": 4.756226856587168e-06, + "loss": 0.0115, + "step": 90280 + }, + { + "epoch": 26.270002909514112, + "grad_norm": 0.08870021253824234, + "learning_rate": 4.753207617652072e-06, + "loss": 0.0123, + "step": 90290 + }, + { + "epoch": 26.272912423625254, + "grad_norm": 0.10023543238639832, + "learning_rate": 4.7501891568932385e-06, + "loss": 0.0108, + "step": 90300 + }, + { + "epoch": 26.2758219377364, + "grad_norm": 0.14649999141693115, + "learning_rate": 4.747171474539891e-06, + "loss": 0.0106, + "step": 90310 + }, + { + "epoch": 26.27873145184754, + "grad_norm": 0.2742539644241333, + "learning_rate": 4.744154570821213e-06, + "loss": 0.0122, + "step": 90320 + }, + { + "epoch": 26.281640965958683, + "grad_norm": 0.1050809994339943, + "learning_rate": 4.741138445966307e-06, + "loss": 0.0123, + "step": 90330 + }, + { + "epoch": 26.28455048006983, + "grad_norm": 0.13148674368858337, + "learning_rate": 4.738123100204238e-06, + "loss": 0.013, + "step": 90340 + }, + { + "epoch": 26.28745999418097, + "grad_norm": 0.5291737914085388, + "learning_rate": 4.7351085337639925e-06, + "loss": 0.0119, + "step": 90350 + }, + { + "epoch": 26.290369508292116, + "grad_norm": 0.15438531339168549, + "learning_rate": 4.732094746874513e-06, + "loss": 0.0117, + "step": 90360 + }, + { + "epoch": 26.293279022403258, + "grad_norm": 0.08039068430662155, + "learning_rate": 4.729081739764678e-06, + "loss": 0.0104, + "step": 90370 + }, + { + "epoch": 26.296188536514403, + "grad_norm": 0.0983627662062645, + "learning_rate": 4.726069512663303e-06, + "loss": 0.0142, + "step": 90380 + }, + { + "epoch": 26.299098050625545, + "grad_norm": 0.08594179898500443, + "learning_rate": 4.723058065799139e-06, + "loss": 0.0114, + "step": 90390 + }, + { + "epoch": 26.302007564736687, + "grad_norm": 0.09103954583406448, + "learning_rate": 4.720047399400899e-06, + "loss": 0.0107, + "step": 90400 + }, + { + "epoch": 26.304917078847833, + "grad_norm": 0.11164502799510956, + "learning_rate": 4.717037513697221e-06, + "loss": 0.0124, + "step": 90410 + }, + { + "epoch": 26.307826592958975, + "grad_norm": 0.12137524783611298, + "learning_rate": 4.714028408916675e-06, + "loss": 0.0139, + "step": 90420 + }, + { + "epoch": 26.31073610707012, + "grad_norm": 0.10045269876718521, + "learning_rate": 4.711020085287791e-06, + "loss": 0.0114, + "step": 90430 + }, + { + "epoch": 26.313645621181262, + "grad_norm": 0.1035187840461731, + "learning_rate": 4.708012543039034e-06, + "loss": 0.0098, + "step": 90440 + }, + { + "epoch": 26.316555135292408, + "grad_norm": 0.0985151007771492, + "learning_rate": 4.705005782398803e-06, + "loss": 0.0115, + "step": 90450 + }, + { + "epoch": 26.31946464940355, + "grad_norm": 0.12011783570051193, + "learning_rate": 4.701999803595437e-06, + "loss": 0.0134, + "step": 90460 + }, + { + "epoch": 26.32237416351469, + "grad_norm": 0.10702847689390182, + "learning_rate": 4.698994606857225e-06, + "loss": 0.0142, + "step": 90470 + }, + { + "epoch": 26.325283677625837, + "grad_norm": 0.09349724650382996, + "learning_rate": 4.695990192412397e-06, + "loss": 0.0136, + "step": 90480 + }, + { + "epoch": 26.32819319173698, + "grad_norm": 0.11306893825531006, + "learning_rate": 4.692986560489108e-06, + "loss": 0.0097, + "step": 90490 + }, + { + "epoch": 26.331102705848124, + "grad_norm": 0.13912679255008698, + "learning_rate": 4.689983711315472e-06, + "loss": 0.0098, + "step": 90500 + }, + { + "epoch": 26.334012219959266, + "grad_norm": 0.1072721928358078, + "learning_rate": 4.6869816451195346e-06, + "loss": 0.0118, + "step": 90510 + }, + { + "epoch": 26.336921734070412, + "grad_norm": 0.07842212170362473, + "learning_rate": 4.683980362129283e-06, + "loss": 0.0105, + "step": 90520 + }, + { + "epoch": 26.339831248181554, + "grad_norm": 0.08809266239404678, + "learning_rate": 4.680979862572639e-06, + "loss": 0.0111, + "step": 90530 + }, + { + "epoch": 26.342740762292696, + "grad_norm": 0.14273081719875336, + "learning_rate": 4.677980146677472e-06, + "loss": 0.0134, + "step": 90540 + }, + { + "epoch": 26.34565027640384, + "grad_norm": 0.11116701364517212, + "learning_rate": 4.6749812146716005e-06, + "loss": 0.0139, + "step": 90550 + }, + { + "epoch": 26.348559790514983, + "grad_norm": 0.09358370304107666, + "learning_rate": 4.6719830667827615e-06, + "loss": 0.0138, + "step": 90560 + }, + { + "epoch": 26.35146930462613, + "grad_norm": 0.1163971871137619, + "learning_rate": 4.668985703238654e-06, + "loss": 0.011, + "step": 90570 + }, + { + "epoch": 26.35437881873727, + "grad_norm": 0.09496762603521347, + "learning_rate": 4.665989124266898e-06, + "loss": 0.0093, + "step": 90580 + }, + { + "epoch": 26.357288332848416, + "grad_norm": 0.1250516027212143, + "learning_rate": 4.662993330095076e-06, + "loss": 0.0122, + "step": 90590 + }, + { + "epoch": 26.360197846959558, + "grad_norm": 0.1159338429570198, + "learning_rate": 4.659998320950685e-06, + "loss": 0.0122, + "step": 90600 + }, + { + "epoch": 26.3631073610707, + "grad_norm": 0.12476833909749985, + "learning_rate": 4.657004097061186e-06, + "loss": 0.0122, + "step": 90610 + }, + { + "epoch": 26.366016875181845, + "grad_norm": 0.10276801139116287, + "learning_rate": 4.6540106586539714e-06, + "loss": 0.012, + "step": 90620 + }, + { + "epoch": 26.368926389292987, + "grad_norm": 0.11106456816196442, + "learning_rate": 4.6510180059563654e-06, + "loss": 0.0117, + "step": 90630 + }, + { + "epoch": 26.371835903404133, + "grad_norm": 0.09830178320407867, + "learning_rate": 4.648026139195649e-06, + "loss": 0.0126, + "step": 90640 + }, + { + "epoch": 26.374745417515275, + "grad_norm": 0.09898657351732254, + "learning_rate": 4.645035058599025e-06, + "loss": 0.0103, + "step": 90650 + }, + { + "epoch": 26.377654931626417, + "grad_norm": 0.09150796383619308, + "learning_rate": 4.6420447643936534e-06, + "loss": 0.0134, + "step": 90660 + }, + { + "epoch": 26.380564445737562, + "grad_norm": 0.11796854436397552, + "learning_rate": 4.639055256806629e-06, + "loss": 0.0113, + "step": 90670 + }, + { + "epoch": 26.383473959848704, + "grad_norm": 0.17374378442764282, + "learning_rate": 4.63606653606498e-06, + "loss": 0.0111, + "step": 90680 + }, + { + "epoch": 26.38638347395985, + "grad_norm": 0.12487521022558212, + "learning_rate": 4.6330786023956786e-06, + "loss": 0.013, + "step": 90690 + }, + { + "epoch": 26.38929298807099, + "grad_norm": 0.12524235248565674, + "learning_rate": 4.630091456025641e-06, + "loss": 0.0119, + "step": 90700 + }, + { + "epoch": 26.392202502182137, + "grad_norm": 0.09337666630744934, + "learning_rate": 4.627105097181728e-06, + "loss": 0.0096, + "step": 90710 + }, + { + "epoch": 26.39511201629328, + "grad_norm": 0.10022281110286713, + "learning_rate": 4.624119526090722e-06, + "loss": 0.0117, + "step": 90720 + }, + { + "epoch": 26.39802153040442, + "grad_norm": 0.11115587502717972, + "learning_rate": 4.621134742979361e-06, + "loss": 0.0119, + "step": 90730 + }, + { + "epoch": 26.400931044515566, + "grad_norm": 0.1378472000360489, + "learning_rate": 4.618150748074329e-06, + "loss": 0.0122, + "step": 90740 + }, + { + "epoch": 26.403840558626708, + "grad_norm": 0.10111185908317566, + "learning_rate": 4.615167541602233e-06, + "loss": 0.0114, + "step": 90750 + }, + { + "epoch": 26.406750072737854, + "grad_norm": 0.10256028920412064, + "learning_rate": 4.6121851237896244e-06, + "loss": 0.0114, + "step": 90760 + }, + { + "epoch": 26.409659586848996, + "grad_norm": 0.08003424108028412, + "learning_rate": 4.609203494863e-06, + "loss": 0.0096, + "step": 90770 + }, + { + "epoch": 26.41256910096014, + "grad_norm": 0.1134411096572876, + "learning_rate": 4.606222655048804e-06, + "loss": 0.0114, + "step": 90780 + }, + { + "epoch": 26.415478615071283, + "grad_norm": 0.08122224360704422, + "learning_rate": 4.603242604573403e-06, + "loss": 0.0111, + "step": 90790 + }, + { + "epoch": 26.418388129182425, + "grad_norm": 0.091563381254673, + "learning_rate": 4.600263343663104e-06, + "loss": 0.0103, + "step": 90800 + }, + { + "epoch": 26.42129764329357, + "grad_norm": 0.12655572593212128, + "learning_rate": 4.597284872544181e-06, + "loss": 0.0106, + "step": 90810 + }, + { + "epoch": 26.424207157404712, + "grad_norm": 0.0899205282330513, + "learning_rate": 4.59430719144282e-06, + "loss": 0.009, + "step": 90820 + }, + { + "epoch": 26.427116671515858, + "grad_norm": 0.08645036071538925, + "learning_rate": 4.591330300585152e-06, + "loss": 0.0133, + "step": 90830 + }, + { + "epoch": 26.430026185627, + "grad_norm": 0.0885079950094223, + "learning_rate": 4.588354200197257e-06, + "loss": 0.0105, + "step": 90840 + }, + { + "epoch": 26.432935699738145, + "grad_norm": 0.3646257817745209, + "learning_rate": 4.585378890505154e-06, + "loss": 0.0097, + "step": 90850 + }, + { + "epoch": 26.435845213849287, + "grad_norm": 0.10318149626255035, + "learning_rate": 4.582404371734787e-06, + "loss": 0.0122, + "step": 90860 + }, + { + "epoch": 26.43875472796043, + "grad_norm": 0.08863361924886703, + "learning_rate": 4.5794306441120645e-06, + "loss": 0.011, + "step": 90870 + }, + { + "epoch": 26.441664242071575, + "grad_norm": 0.09441103786230087, + "learning_rate": 4.57645770786281e-06, + "loss": 0.0113, + "step": 90880 + }, + { + "epoch": 26.444573756182717, + "grad_norm": 0.1265801042318344, + "learning_rate": 4.573485563212807e-06, + "loss": 0.0115, + "step": 90890 + }, + { + "epoch": 26.447483270293862, + "grad_norm": 0.10067295283079147, + "learning_rate": 4.5705142103877626e-06, + "loss": 0.0109, + "step": 90900 + }, + { + "epoch": 26.450392784405004, + "grad_norm": 0.18609189987182617, + "learning_rate": 4.567543649613336e-06, + "loss": 0.0117, + "step": 90910 + }, + { + "epoch": 26.45330229851615, + "grad_norm": 0.10647151619195938, + "learning_rate": 4.564573881115124e-06, + "loss": 0.0108, + "step": 90920 + }, + { + "epoch": 26.45621181262729, + "grad_norm": 0.08267899602651596, + "learning_rate": 4.561604905118654e-06, + "loss": 0.0116, + "step": 90930 + }, + { + "epoch": 26.459121326738433, + "grad_norm": 0.09946750104427338, + "learning_rate": 4.558636721849409e-06, + "loss": 0.0132, + "step": 90940 + }, + { + "epoch": 26.46203084084958, + "grad_norm": 0.10589633136987686, + "learning_rate": 4.555669331532793e-06, + "loss": 0.0142, + "step": 90950 + }, + { + "epoch": 26.46494035496072, + "grad_norm": 0.10588102787733078, + "learning_rate": 4.5527027343941655e-06, + "loss": 0.0135, + "step": 90960 + }, + { + "epoch": 26.467849869071866, + "grad_norm": 0.22374524176120758, + "learning_rate": 4.549736930658823e-06, + "loss": 0.0112, + "step": 90970 + }, + { + "epoch": 26.470759383183008, + "grad_norm": 0.12799575924873352, + "learning_rate": 4.5467719205519945e-06, + "loss": 0.0129, + "step": 90980 + }, + { + "epoch": 26.473668897294154, + "grad_norm": 0.1023198589682579, + "learning_rate": 4.543807704298849e-06, + "loss": 0.013, + "step": 90990 + }, + { + "epoch": 26.476578411405296, + "grad_norm": 0.1371038556098938, + "learning_rate": 4.540844282124504e-06, + "loss": 0.0115, + "step": 91000 + }, + { + "epoch": 26.479487925516437, + "grad_norm": 0.0804385393857956, + "learning_rate": 4.537881654254015e-06, + "loss": 0.0125, + "step": 91010 + }, + { + "epoch": 26.482397439627583, + "grad_norm": 0.0879087895154953, + "learning_rate": 4.534919820912366e-06, + "loss": 0.0102, + "step": 91020 + }, + { + "epoch": 26.485306953738725, + "grad_norm": 0.12771983444690704, + "learning_rate": 4.531958782324495e-06, + "loss": 0.0134, + "step": 91030 + }, + { + "epoch": 26.48821646784987, + "grad_norm": 0.09454658627510071, + "learning_rate": 4.528998538715274e-06, + "loss": 0.0108, + "step": 91040 + }, + { + "epoch": 26.491125981961012, + "grad_norm": 0.07814504206180573, + "learning_rate": 4.526039090309512e-06, + "loss": 0.0109, + "step": 91050 + }, + { + "epoch": 26.494035496072154, + "grad_norm": 0.1336139738559723, + "learning_rate": 4.523080437331953e-06, + "loss": 0.0133, + "step": 91060 + }, + { + "epoch": 26.4969450101833, + "grad_norm": 0.16186927258968353, + "learning_rate": 4.520122580007295e-06, + "loss": 0.0134, + "step": 91070 + }, + { + "epoch": 26.49985452429444, + "grad_norm": 0.13468362390995026, + "learning_rate": 4.51716551856017e-06, + "loss": 0.0122, + "step": 91080 + }, + { + "epoch": 26.502764038405587, + "grad_norm": 0.08516331762075424, + "learning_rate": 4.514209253215143e-06, + "loss": 0.0102, + "step": 91090 + }, + { + "epoch": 26.50567355251673, + "grad_norm": 0.1115460991859436, + "learning_rate": 4.511253784196714e-06, + "loss": 0.0126, + "step": 91100 + }, + { + "epoch": 26.508583066627875, + "grad_norm": 0.10566014796495438, + "learning_rate": 4.50829911172935e-06, + "loss": 0.0092, + "step": 91110 + }, + { + "epoch": 26.511492580739016, + "grad_norm": 0.09678252786397934, + "learning_rate": 4.505345236037428e-06, + "loss": 0.0092, + "step": 91120 + }, + { + "epoch": 26.51440209485016, + "grad_norm": 0.12016180902719498, + "learning_rate": 4.502392157345272e-06, + "loss": 0.0107, + "step": 91130 + }, + { + "epoch": 26.517311608961304, + "grad_norm": 0.12916220724582672, + "learning_rate": 4.4994398758771555e-06, + "loss": 0.0106, + "step": 91140 + }, + { + "epoch": 26.520221123072446, + "grad_norm": 0.09976735711097717, + "learning_rate": 4.496488391857285e-06, + "loss": 0.0134, + "step": 91150 + }, + { + "epoch": 26.52313063718359, + "grad_norm": 0.11443876475095749, + "learning_rate": 4.4935377055098055e-06, + "loss": 0.0111, + "step": 91160 + }, + { + "epoch": 26.526040151294733, + "grad_norm": 0.13520805537700653, + "learning_rate": 4.490587817058795e-06, + "loss": 0.0104, + "step": 91170 + }, + { + "epoch": 26.52894966540588, + "grad_norm": 0.1370030790567398, + "learning_rate": 4.487638726728285e-06, + "loss": 0.0131, + "step": 91180 + }, + { + "epoch": 26.53185917951702, + "grad_norm": 0.09787514805793762, + "learning_rate": 4.484690434742242e-06, + "loss": 0.0104, + "step": 91190 + }, + { + "epoch": 26.534768693628163, + "grad_norm": 0.08659069240093231, + "learning_rate": 4.481742941324561e-06, + "loss": 0.0098, + "step": 91200 + }, + { + "epoch": 26.537678207739308, + "grad_norm": 0.1023460105061531, + "learning_rate": 4.478796246699088e-06, + "loss": 0.0113, + "step": 91210 + }, + { + "epoch": 26.54058772185045, + "grad_norm": 0.07952608168125153, + "learning_rate": 4.475850351089613e-06, + "loss": 0.0092, + "step": 91220 + }, + { + "epoch": 26.543497235961595, + "grad_norm": 0.2599896192550659, + "learning_rate": 4.472905254719844e-06, + "loss": 0.0107, + "step": 91230 + }, + { + "epoch": 26.546406750072737, + "grad_norm": 0.11218587309122086, + "learning_rate": 4.469960957813453e-06, + "loss": 0.0117, + "step": 91240 + }, + { + "epoch": 26.549316264183883, + "grad_norm": 0.14861713349819183, + "learning_rate": 4.467017460594032e-06, + "loss": 0.0117, + "step": 91250 + }, + { + "epoch": 26.552225778295025, + "grad_norm": 0.17651528120040894, + "learning_rate": 4.464074763285126e-06, + "loss": 0.0102, + "step": 91260 + }, + { + "epoch": 26.555135292406167, + "grad_norm": 0.09837374836206436, + "learning_rate": 4.461132866110207e-06, + "loss": 0.0094, + "step": 91270 + }, + { + "epoch": 26.558044806517312, + "grad_norm": 0.09863891452550888, + "learning_rate": 4.458191769292701e-06, + "loss": 0.0119, + "step": 91280 + }, + { + "epoch": 26.560954320628454, + "grad_norm": 0.12719625234603882, + "learning_rate": 4.455251473055955e-06, + "loss": 0.011, + "step": 91290 + }, + { + "epoch": 26.5638638347396, + "grad_norm": 0.1400635540485382, + "learning_rate": 4.452311977623271e-06, + "loss": 0.0129, + "step": 91300 + }, + { + "epoch": 26.56677334885074, + "grad_norm": 0.1048617735505104, + "learning_rate": 4.449373283217887e-06, + "loss": 0.0109, + "step": 91310 + }, + { + "epoch": 26.569682862961887, + "grad_norm": 0.10718239098787308, + "learning_rate": 4.446435390062971e-06, + "loss": 0.0103, + "step": 91320 + }, + { + "epoch": 26.57259237707303, + "grad_norm": 0.11520854383707047, + "learning_rate": 4.4434982983816394e-06, + "loss": 0.0125, + "step": 91330 + }, + { + "epoch": 26.57550189118417, + "grad_norm": 0.11858239024877548, + "learning_rate": 4.44056200839695e-06, + "loss": 0.0101, + "step": 91340 + }, + { + "epoch": 26.578411405295316, + "grad_norm": 0.09878456592559814, + "learning_rate": 4.437626520331889e-06, + "loss": 0.0111, + "step": 91350 + }, + { + "epoch": 26.58132091940646, + "grad_norm": 0.12802881002426147, + "learning_rate": 4.434691834409384e-06, + "loss": 0.0115, + "step": 91360 + }, + { + "epoch": 26.584230433517604, + "grad_norm": 0.07962270826101303, + "learning_rate": 4.431757950852308e-06, + "loss": 0.0099, + "step": 91370 + }, + { + "epoch": 26.587139947628746, + "grad_norm": 0.08883122354745865, + "learning_rate": 4.428824869883478e-06, + "loss": 0.0129, + "step": 91380 + }, + { + "epoch": 26.590049461739888, + "grad_norm": 0.08238878101110458, + "learning_rate": 4.425892591725635e-06, + "loss": 0.0111, + "step": 91390 + }, + { + "epoch": 26.592958975851033, + "grad_norm": 0.11293130367994308, + "learning_rate": 4.422961116601457e-06, + "loss": 0.0113, + "step": 91400 + }, + { + "epoch": 26.595868489962175, + "grad_norm": 0.08979939669370651, + "learning_rate": 4.420030444733587e-06, + "loss": 0.0124, + "step": 91410 + }, + { + "epoch": 26.59877800407332, + "grad_norm": 0.12223154306411743, + "learning_rate": 4.417100576344586e-06, + "loss": 0.011, + "step": 91420 + }, + { + "epoch": 26.601687518184463, + "grad_norm": 0.09921611845493317, + "learning_rate": 4.414171511656948e-06, + "loss": 0.0114, + "step": 91430 + }, + { + "epoch": 26.604597032295608, + "grad_norm": 0.1262313723564148, + "learning_rate": 4.411243250893125e-06, + "loss": 0.0154, + "step": 91440 + }, + { + "epoch": 26.60750654640675, + "grad_norm": 0.09721556305885315, + "learning_rate": 4.408315794275502e-06, + "loss": 0.0106, + "step": 91450 + }, + { + "epoch": 26.610416060517892, + "grad_norm": 0.09846573323011398, + "learning_rate": 4.405389142026395e-06, + "loss": 0.0112, + "step": 91460 + }, + { + "epoch": 26.613325574629037, + "grad_norm": 0.11640197038650513, + "learning_rate": 4.402463294368056e-06, + "loss": 0.0124, + "step": 91470 + }, + { + "epoch": 26.61623508874018, + "grad_norm": 0.14038334786891937, + "learning_rate": 4.399538251522701e-06, + "loss": 0.01, + "step": 91480 + }, + { + "epoch": 26.619144602851325, + "grad_norm": 0.08407562226057053, + "learning_rate": 4.396614013712458e-06, + "loss": 0.0124, + "step": 91490 + }, + { + "epoch": 26.622054116962467, + "grad_norm": 0.08696401119232178, + "learning_rate": 4.393690581159399e-06, + "loss": 0.0118, + "step": 91500 + }, + { + "epoch": 26.624963631073612, + "grad_norm": 0.10573653131723404, + "learning_rate": 4.390767954085548e-06, + "loss": 0.0128, + "step": 91510 + }, + { + "epoch": 26.627873145184754, + "grad_norm": 0.09253575652837753, + "learning_rate": 4.3878461327128585e-06, + "loss": 0.0129, + "step": 91520 + }, + { + "epoch": 26.630782659295896, + "grad_norm": 0.09447692334651947, + "learning_rate": 4.38492511726322e-06, + "loss": 0.0114, + "step": 91530 + }, + { + "epoch": 26.63369217340704, + "grad_norm": 0.12340085208415985, + "learning_rate": 4.382004907958462e-06, + "loss": 0.0114, + "step": 91540 + }, + { + "epoch": 26.636601687518183, + "grad_norm": 0.1250937134027481, + "learning_rate": 4.37908550502036e-06, + "loss": 0.0101, + "step": 91550 + }, + { + "epoch": 26.63951120162933, + "grad_norm": 0.10686202347278595, + "learning_rate": 4.376166908670624e-06, + "loss": 0.0124, + "step": 91560 + }, + { + "epoch": 26.64242071574047, + "grad_norm": 0.13538463413715363, + "learning_rate": 4.373249119130897e-06, + "loss": 0.0106, + "step": 91570 + }, + { + "epoch": 26.645330229851616, + "grad_norm": 0.1337747424840927, + "learning_rate": 4.370332136622768e-06, + "loss": 0.0111, + "step": 91580 + }, + { + "epoch": 26.64823974396276, + "grad_norm": 0.09373465925455093, + "learning_rate": 4.367415961367769e-06, + "loss": 0.0113, + "step": 91590 + }, + { + "epoch": 26.6511492580739, + "grad_norm": 0.09710672497749329, + "learning_rate": 4.364500593587355e-06, + "loss": 0.0114, + "step": 91600 + }, + { + "epoch": 26.654058772185046, + "grad_norm": 0.12030233442783356, + "learning_rate": 4.361586033502936e-06, + "loss": 0.0103, + "step": 91610 + }, + { + "epoch": 26.656968286296188, + "grad_norm": 0.08214841037988663, + "learning_rate": 4.358672281335845e-06, + "loss": 0.0108, + "step": 91620 + }, + { + "epoch": 26.659877800407333, + "grad_norm": 0.08227057009935379, + "learning_rate": 4.355759337307374e-06, + "loss": 0.0104, + "step": 91630 + }, + { + "epoch": 26.662787314518475, + "grad_norm": 0.0918714627623558, + "learning_rate": 4.35284720163873e-06, + "loss": 0.0114, + "step": 91640 + }, + { + "epoch": 26.66569682862962, + "grad_norm": 0.16049595177173615, + "learning_rate": 4.34993587455108e-06, + "loss": 0.0133, + "step": 91650 + }, + { + "epoch": 26.668606342740762, + "grad_norm": 0.09874666482210159, + "learning_rate": 4.347025356265514e-06, + "loss": 0.0122, + "step": 91660 + }, + { + "epoch": 26.671515856851904, + "grad_norm": 0.08409792184829712, + "learning_rate": 4.3441156470030656e-06, + "loss": 0.0121, + "step": 91670 + }, + { + "epoch": 26.67442537096305, + "grad_norm": 0.09902283549308777, + "learning_rate": 4.341206746984717e-06, + "loss": 0.0112, + "step": 91680 + }, + { + "epoch": 26.677334885074192, + "grad_norm": 0.10682518780231476, + "learning_rate": 4.338298656431369e-06, + "loss": 0.0111, + "step": 91690 + }, + { + "epoch": 26.680244399185337, + "grad_norm": 0.09408320486545563, + "learning_rate": 4.335391375563876e-06, + "loss": 0.0108, + "step": 91700 + }, + { + "epoch": 26.68315391329648, + "grad_norm": 0.19932375848293304, + "learning_rate": 4.332484904603033e-06, + "loss": 0.0115, + "step": 91710 + }, + { + "epoch": 26.686063427407625, + "grad_norm": 0.2168942242860794, + "learning_rate": 4.329579243769564e-06, + "loss": 0.0116, + "step": 91720 + }, + { + "epoch": 26.688972941518767, + "grad_norm": 0.12121767550706863, + "learning_rate": 4.326674393284127e-06, + "loss": 0.0123, + "step": 91730 + }, + { + "epoch": 26.69188245562991, + "grad_norm": 0.13162629306316376, + "learning_rate": 4.3237703533673314e-06, + "loss": 0.0116, + "step": 91740 + }, + { + "epoch": 26.694791969741054, + "grad_norm": 0.13065341114997864, + "learning_rate": 4.320867124239726e-06, + "loss": 0.0111, + "step": 91750 + }, + { + "epoch": 26.697701483852196, + "grad_norm": 0.15767158567905426, + "learning_rate": 4.317964706121786e-06, + "loss": 0.0104, + "step": 91760 + }, + { + "epoch": 26.70061099796334, + "grad_norm": 0.12849470973014832, + "learning_rate": 4.3150630992339205e-06, + "loss": 0.0108, + "step": 91770 + }, + { + "epoch": 26.703520512074483, + "grad_norm": 0.11732284724712372, + "learning_rate": 4.312162303796509e-06, + "loss": 0.0119, + "step": 91780 + }, + { + "epoch": 26.706430026185625, + "grad_norm": 0.13909600675106049, + "learning_rate": 4.309262320029837e-06, + "loss": 0.0135, + "step": 91790 + }, + { + "epoch": 26.70933954029677, + "grad_norm": 0.1262485831975937, + "learning_rate": 4.3063631481541334e-06, + "loss": 0.0107, + "step": 91800 + }, + { + "epoch": 26.712249054407913, + "grad_norm": 0.09967502951622009, + "learning_rate": 4.3034647883895785e-06, + "loss": 0.0115, + "step": 91810 + }, + { + "epoch": 26.71515856851906, + "grad_norm": 0.0806422233581543, + "learning_rate": 4.3005672409562855e-06, + "loss": 0.0129, + "step": 91820 + }, + { + "epoch": 26.7180680826302, + "grad_norm": 0.11641861498355865, + "learning_rate": 4.297670506074299e-06, + "loss": 0.012, + "step": 91830 + }, + { + "epoch": 26.720977596741346, + "grad_norm": 0.10879762470722198, + "learning_rate": 4.2947745839636065e-06, + "loss": 0.0107, + "step": 91840 + }, + { + "epoch": 26.723887110852488, + "grad_norm": 0.12316453456878662, + "learning_rate": 4.291879474844135e-06, + "loss": 0.0101, + "step": 91850 + }, + { + "epoch": 26.72679662496363, + "grad_norm": 0.12228534370660782, + "learning_rate": 4.288985178935753e-06, + "loss": 0.0125, + "step": 91860 + }, + { + "epoch": 26.729706139074775, + "grad_norm": 0.06362485140562057, + "learning_rate": 4.286091696458257e-06, + "loss": 0.0098, + "step": 91870 + }, + { + "epoch": 26.732615653185917, + "grad_norm": 0.1321079283952713, + "learning_rate": 4.283199027631392e-06, + "loss": 0.0131, + "step": 91880 + }, + { + "epoch": 26.735525167297062, + "grad_norm": 0.10637672245502472, + "learning_rate": 4.280307172674839e-06, + "loss": 0.0114, + "step": 91890 + }, + { + "epoch": 26.738434681408204, + "grad_norm": 0.08570706844329834, + "learning_rate": 4.277416131808212e-06, + "loss": 0.0117, + "step": 91900 + }, + { + "epoch": 26.74134419551935, + "grad_norm": 0.10658394545316696, + "learning_rate": 4.2745259052510645e-06, + "loss": 0.0107, + "step": 91910 + }, + { + "epoch": 26.74425370963049, + "grad_norm": 0.06176143139600754, + "learning_rate": 4.27163649322289e-06, + "loss": 0.0114, + "step": 91920 + }, + { + "epoch": 26.747163223741634, + "grad_norm": 0.08122782409191132, + "learning_rate": 4.268747895943129e-06, + "loss": 0.0113, + "step": 91930 + }, + { + "epoch": 26.75007273785278, + "grad_norm": 0.10097743570804596, + "learning_rate": 4.265860113631142e-06, + "loss": 0.0105, + "step": 91940 + }, + { + "epoch": 26.75298225196392, + "grad_norm": 0.09584463387727737, + "learning_rate": 4.262973146506241e-06, + "loss": 0.0117, + "step": 91950 + }, + { + "epoch": 26.755891766075067, + "grad_norm": 0.11343914270401001, + "learning_rate": 4.260086994787669e-06, + "loss": 0.0111, + "step": 91960 + }, + { + "epoch": 26.75880128018621, + "grad_norm": 0.12820829451084137, + "learning_rate": 4.257201658694612e-06, + "loss": 0.0135, + "step": 91970 + }, + { + "epoch": 26.761710794297354, + "grad_norm": 0.1539735049009323, + "learning_rate": 4.2543171384461985e-06, + "loss": 0.0143, + "step": 91980 + }, + { + "epoch": 26.764620308408496, + "grad_norm": 0.11787393689155579, + "learning_rate": 4.251433434261477e-06, + "loss": 0.0106, + "step": 91990 + }, + { + "epoch": 26.767529822519638, + "grad_norm": 0.10539472848176956, + "learning_rate": 4.248550546359458e-06, + "loss": 0.011, + "step": 92000 + }, + { + "epoch": 26.770439336630783, + "grad_norm": 0.13080309331417084, + "learning_rate": 4.245668474959065e-06, + "loss": 0.0142, + "step": 92010 + }, + { + "epoch": 26.773348850741925, + "grad_norm": 0.26983627676963806, + "learning_rate": 4.242787220279185e-06, + "loss": 0.0125, + "step": 92020 + }, + { + "epoch": 26.77625836485307, + "grad_norm": 0.09890328347682953, + "learning_rate": 4.23990678253862e-06, + "loss": 0.0116, + "step": 92030 + }, + { + "epoch": 26.779167878964213, + "grad_norm": 0.11391404271125793, + "learning_rate": 4.237027161956122e-06, + "loss": 0.0117, + "step": 92040 + }, + { + "epoch": 26.782077393075355, + "grad_norm": 0.08783342689275742, + "learning_rate": 4.234148358750388e-06, + "loss": 0.0114, + "step": 92050 + }, + { + "epoch": 26.7849869071865, + "grad_norm": 0.1373830884695053, + "learning_rate": 4.231270373140037e-06, + "loss": 0.012, + "step": 92060 + }, + { + "epoch": 26.787896421297642, + "grad_norm": 0.11733800172805786, + "learning_rate": 4.228393205343626e-06, + "loss": 0.0135, + "step": 92070 + }, + { + "epoch": 26.790805935408788, + "grad_norm": 0.10977676510810852, + "learning_rate": 4.2255168555796715e-06, + "loss": 0.012, + "step": 92080 + }, + { + "epoch": 26.79371544951993, + "grad_norm": 0.08962979167699814, + "learning_rate": 4.2226413240666086e-06, + "loss": 0.011, + "step": 92090 + }, + { + "epoch": 26.796624963631075, + "grad_norm": 0.08938459306955338, + "learning_rate": 4.219766611022807e-06, + "loss": 0.0118, + "step": 92100 + }, + { + "epoch": 26.799534477742217, + "grad_norm": 0.0984835997223854, + "learning_rate": 4.216892716666588e-06, + "loss": 0.01, + "step": 92110 + }, + { + "epoch": 26.80244399185336, + "grad_norm": 0.10502730309963226, + "learning_rate": 4.214019641216209e-06, + "loss": 0.0118, + "step": 92120 + }, + { + "epoch": 26.805353505964504, + "grad_norm": 0.08048006147146225, + "learning_rate": 4.211147384889858e-06, + "loss": 0.0101, + "step": 92130 + }, + { + "epoch": 26.808263020075646, + "grad_norm": 0.11698340624570847, + "learning_rate": 4.208275947905659e-06, + "loss": 0.0123, + "step": 92140 + }, + { + "epoch": 26.81117253418679, + "grad_norm": 0.10749784857034683, + "learning_rate": 4.205405330481681e-06, + "loss": 0.0127, + "step": 92150 + }, + { + "epoch": 26.814082048297934, + "grad_norm": 0.12078143656253815, + "learning_rate": 4.202535532835934e-06, + "loss": 0.011, + "step": 92160 + }, + { + "epoch": 26.81699156240908, + "grad_norm": 0.11562138795852661, + "learning_rate": 4.199666555186354e-06, + "loss": 0.0127, + "step": 92170 + }, + { + "epoch": 26.81990107652022, + "grad_norm": 0.09818000346422195, + "learning_rate": 4.196798397750821e-06, + "loss": 0.0122, + "step": 92180 + }, + { + "epoch": 26.822810590631363, + "grad_norm": 0.12329849600791931, + "learning_rate": 4.1939310607471595e-06, + "loss": 0.0121, + "step": 92190 + }, + { + "epoch": 26.82572010474251, + "grad_norm": 0.08824770897626877, + "learning_rate": 4.191064544393122e-06, + "loss": 0.0144, + "step": 92200 + }, + { + "epoch": 26.82862961885365, + "grad_norm": 0.08534722030162811, + "learning_rate": 4.188198848906393e-06, + "loss": 0.0101, + "step": 92210 + }, + { + "epoch": 26.831539132964796, + "grad_norm": 0.14103595912456512, + "learning_rate": 4.185333974504611e-06, + "loss": 0.0112, + "step": 92220 + }, + { + "epoch": 26.834448647075938, + "grad_norm": 0.09276002645492554, + "learning_rate": 4.1824699214053474e-06, + "loss": 0.0102, + "step": 92230 + }, + { + "epoch": 26.837358161187083, + "grad_norm": 0.13936731219291687, + "learning_rate": 4.179606689826098e-06, + "loss": 0.0115, + "step": 92240 + }, + { + "epoch": 26.840267675298225, + "grad_norm": 0.12857003509998322, + "learning_rate": 4.176744279984319e-06, + "loss": 0.0123, + "step": 92250 + }, + { + "epoch": 26.843177189409367, + "grad_norm": 0.10333877801895142, + "learning_rate": 4.173882692097379e-06, + "loss": 0.0108, + "step": 92260 + }, + { + "epoch": 26.846086703520513, + "grad_norm": 0.08836212754249573, + "learning_rate": 4.171021926382605e-06, + "loss": 0.0095, + "step": 92270 + }, + { + "epoch": 26.848996217631655, + "grad_norm": 0.10284505039453506, + "learning_rate": 4.1681619830572484e-06, + "loss": 0.0116, + "step": 92280 + }, + { + "epoch": 26.8519057317428, + "grad_norm": 0.11580946296453476, + "learning_rate": 4.165302862338504e-06, + "loss": 0.0098, + "step": 92290 + }, + { + "epoch": 26.854815245853942, + "grad_norm": 0.1068500280380249, + "learning_rate": 4.1624445644435094e-06, + "loss": 0.0145, + "step": 92300 + }, + { + "epoch": 26.857724759965087, + "grad_norm": 0.473379909992218, + "learning_rate": 4.159587089589325e-06, + "loss": 0.0108, + "step": 92310 + }, + { + "epoch": 26.86063427407623, + "grad_norm": 0.11877524852752686, + "learning_rate": 4.156730437992965e-06, + "loss": 0.0114, + "step": 92320 + }, + { + "epoch": 26.86354378818737, + "grad_norm": 0.10467761754989624, + "learning_rate": 4.153874609871364e-06, + "loss": 0.0109, + "step": 92330 + }, + { + "epoch": 26.866453302298517, + "grad_norm": 0.13698382675647736, + "learning_rate": 4.151019605441406e-06, + "loss": 0.0119, + "step": 92340 + }, + { + "epoch": 26.86936281640966, + "grad_norm": 0.095960333943367, + "learning_rate": 4.148165424919919e-06, + "loss": 0.0122, + "step": 92350 + }, + { + "epoch": 26.872272330520804, + "grad_norm": 0.22270652651786804, + "learning_rate": 4.145312068523652e-06, + "loss": 0.011, + "step": 92360 + }, + { + "epoch": 26.875181844631946, + "grad_norm": 0.07740391790866852, + "learning_rate": 4.142459536469292e-06, + "loss": 0.0108, + "step": 92370 + }, + { + "epoch": 26.87809135874309, + "grad_norm": 0.07714121043682098, + "learning_rate": 4.139607828973479e-06, + "loss": 0.0123, + "step": 92380 + }, + { + "epoch": 26.881000872854234, + "grad_norm": 0.1104799434542656, + "learning_rate": 4.136756946252784e-06, + "loss": 0.0116, + "step": 92390 + }, + { + "epoch": 26.883910386965375, + "grad_norm": 0.1066170334815979, + "learning_rate": 4.133906888523702e-06, + "loss": 0.0121, + "step": 92400 + }, + { + "epoch": 26.88681990107652, + "grad_norm": 0.09247691184282303, + "learning_rate": 4.1310576560026816e-06, + "loss": 0.0117, + "step": 92410 + }, + { + "epoch": 26.889729415187663, + "grad_norm": 0.09868945181369781, + "learning_rate": 4.128209248906108e-06, + "loss": 0.0123, + "step": 92420 + }, + { + "epoch": 26.89263892929881, + "grad_norm": 0.10409216582775116, + "learning_rate": 4.125361667450295e-06, + "loss": 0.013, + "step": 92430 + }, + { + "epoch": 26.89554844340995, + "grad_norm": 0.09770956635475159, + "learning_rate": 4.122514911851491e-06, + "loss": 0.0138, + "step": 92440 + }, + { + "epoch": 26.898457957521096, + "grad_norm": 0.10641931742429733, + "learning_rate": 4.119668982325897e-06, + "loss": 0.0118, + "step": 92450 + }, + { + "epoch": 26.901367471632238, + "grad_norm": 0.11176233738660812, + "learning_rate": 4.116823879089642e-06, + "loss": 0.0097, + "step": 92460 + }, + { + "epoch": 26.90427698574338, + "grad_norm": 0.1121622622013092, + "learning_rate": 4.113979602358792e-06, + "loss": 0.0123, + "step": 92470 + }, + { + "epoch": 26.907186499854525, + "grad_norm": 0.10391731560230255, + "learning_rate": 4.11113615234934e-06, + "loss": 0.011, + "step": 92480 + }, + { + "epoch": 26.910096013965667, + "grad_norm": 0.08893290162086487, + "learning_rate": 4.108293529277245e-06, + "loss": 0.0109, + "step": 92490 + }, + { + "epoch": 26.913005528076813, + "grad_norm": 0.1407746821641922, + "learning_rate": 4.105451733358379e-06, + "loss": 0.0113, + "step": 92500 + }, + { + "epoch": 26.915915042187954, + "grad_norm": 0.09385742247104645, + "learning_rate": 4.102610764808551e-06, + "loss": 0.0108, + "step": 92510 + }, + { + "epoch": 26.918824556299096, + "grad_norm": 0.1335674226284027, + "learning_rate": 4.099770623843519e-06, + "loss": 0.0107, + "step": 92520 + }, + { + "epoch": 26.921734070410242, + "grad_norm": 0.07395434379577637, + "learning_rate": 4.096931310678977e-06, + "loss": 0.0112, + "step": 92530 + }, + { + "epoch": 26.924643584521384, + "grad_norm": 0.11884579062461853, + "learning_rate": 4.094092825530544e-06, + "loss": 0.0135, + "step": 92540 + }, + { + "epoch": 26.92755309863253, + "grad_norm": 0.08266336470842361, + "learning_rate": 4.091255168613792e-06, + "loss": 0.0122, + "step": 92550 + }, + { + "epoch": 26.93046261274367, + "grad_norm": 0.10173249989748001, + "learning_rate": 4.0884183401442155e-06, + "loss": 0.0117, + "step": 92560 + }, + { + "epoch": 26.933372126854817, + "grad_norm": 0.1460864245891571, + "learning_rate": 4.08558234033726e-06, + "loss": 0.012, + "step": 92570 + }, + { + "epoch": 26.93628164096596, + "grad_norm": 0.07080257683992386, + "learning_rate": 4.082747169408292e-06, + "loss": 0.0101, + "step": 92580 + }, + { + "epoch": 26.9391911550771, + "grad_norm": 0.10333485156297684, + "learning_rate": 4.079912827572629e-06, + "loss": 0.0131, + "step": 92590 + }, + { + "epoch": 26.942100669188246, + "grad_norm": 0.0816969946026802, + "learning_rate": 4.077079315045524e-06, + "loss": 0.0091, + "step": 92600 + }, + { + "epoch": 26.945010183299388, + "grad_norm": 0.15164759755134583, + "learning_rate": 4.0742466320421565e-06, + "loss": 0.012, + "step": 92610 + }, + { + "epoch": 26.947919697410534, + "grad_norm": 0.1259452998638153, + "learning_rate": 4.071414778777657e-06, + "loss": 0.0098, + "step": 92620 + }, + { + "epoch": 26.950829211521675, + "grad_norm": 0.11392171680927277, + "learning_rate": 4.06858375546708e-06, + "loss": 0.0108, + "step": 92630 + }, + { + "epoch": 26.95373872563282, + "grad_norm": 0.06921622902154922, + "learning_rate": 4.065753562325428e-06, + "loss": 0.0107, + "step": 92640 + }, + { + "epoch": 26.956648239743963, + "grad_norm": 0.19174450635910034, + "learning_rate": 4.062924199567631e-06, + "loss": 0.0121, + "step": 92650 + }, + { + "epoch": 26.959557753855105, + "grad_norm": 0.12076525390148163, + "learning_rate": 4.060095667408566e-06, + "loss": 0.0114, + "step": 92660 + }, + { + "epoch": 26.96246726796625, + "grad_norm": 0.10767622292041779, + "learning_rate": 4.057267966063034e-06, + "loss": 0.0125, + "step": 92670 + }, + { + "epoch": 26.965376782077392, + "grad_norm": 0.15069574117660522, + "learning_rate": 4.054441095745782e-06, + "loss": 0.012, + "step": 92680 + }, + { + "epoch": 26.968286296188538, + "grad_norm": 0.14754079282283783, + "learning_rate": 4.0516150566715005e-06, + "loss": 0.0115, + "step": 92690 + }, + { + "epoch": 26.97119581029968, + "grad_norm": 0.09692368656396866, + "learning_rate": 4.048789849054798e-06, + "loss": 0.0118, + "step": 92700 + }, + { + "epoch": 26.974105324410825, + "grad_norm": 0.10250865668058395, + "learning_rate": 4.045965473110235e-06, + "loss": 0.0125, + "step": 92710 + }, + { + "epoch": 26.977014838521967, + "grad_norm": 0.1325760781764984, + "learning_rate": 4.043141929052309e-06, + "loss": 0.0132, + "step": 92720 + }, + { + "epoch": 26.97992435263311, + "grad_norm": 0.0843382179737091, + "learning_rate": 4.0403192170954435e-06, + "loss": 0.0134, + "step": 92730 + }, + { + "epoch": 26.982833866744254, + "grad_norm": 0.11270768195390701, + "learning_rate": 4.037497337454001e-06, + "loss": 0.0115, + "step": 92740 + }, + { + "epoch": 26.985743380855396, + "grad_norm": 0.18487466871738434, + "learning_rate": 4.034676290342289e-06, + "loss": 0.0108, + "step": 92750 + }, + { + "epoch": 26.988652894966542, + "grad_norm": 0.10621064901351929, + "learning_rate": 4.031856075974555e-06, + "loss": 0.0118, + "step": 92760 + }, + { + "epoch": 26.991562409077684, + "grad_norm": 0.10340508818626404, + "learning_rate": 4.029036694564968e-06, + "loss": 0.0115, + "step": 92770 + }, + { + "epoch": 26.994471923188826, + "grad_norm": 0.08859309554100037, + "learning_rate": 4.0262181463276316e-06, + "loss": 0.0097, + "step": 92780 + }, + { + "epoch": 26.99738143729997, + "grad_norm": 0.1010841578245163, + "learning_rate": 4.023400431476615e-06, + "loss": 0.0123, + "step": 92790 + }, + { + "epoch": 27.000290951411113, + "grad_norm": 0.0801166519522667, + "learning_rate": 4.020583550225896e-06, + "loss": 0.013, + "step": 92800 + }, + { + "epoch": 27.00320046552226, + "grad_norm": 0.17988455295562744, + "learning_rate": 4.0177675027893965e-06, + "loss": 0.0114, + "step": 92810 + }, + { + "epoch": 27.0061099796334, + "grad_norm": 0.12971754372119904, + "learning_rate": 4.014952289380977e-06, + "loss": 0.0122, + "step": 92820 + }, + { + "epoch": 27.009019493744546, + "grad_norm": 0.09873839467763901, + "learning_rate": 4.012137910214441e-06, + "loss": 0.0106, + "step": 92830 + }, + { + "epoch": 27.011929007855688, + "grad_norm": 0.08898330479860306, + "learning_rate": 4.009324365503516e-06, + "loss": 0.0107, + "step": 92840 + }, + { + "epoch": 27.01483852196683, + "grad_norm": 0.1646529585123062, + "learning_rate": 4.006511655461871e-06, + "loss": 0.0122, + "step": 92850 + }, + { + "epoch": 27.017748036077975, + "grad_norm": 0.09366992861032486, + "learning_rate": 4.003699780303114e-06, + "loss": 0.0133, + "step": 92860 + }, + { + "epoch": 27.020657550189117, + "grad_norm": 0.08683478832244873, + "learning_rate": 4.000888740240793e-06, + "loss": 0.011, + "step": 92870 + }, + { + "epoch": 27.023567064300263, + "grad_norm": 0.29735320806503296, + "learning_rate": 3.998078535488382e-06, + "loss": 0.0117, + "step": 92880 + }, + { + "epoch": 27.026476578411405, + "grad_norm": 0.1347411721944809, + "learning_rate": 3.9952691662592995e-06, + "loss": 0.0114, + "step": 92890 + }, + { + "epoch": 27.02938609252255, + "grad_norm": 0.08631578832864761, + "learning_rate": 3.992460632766902e-06, + "loss": 0.0111, + "step": 92900 + }, + { + "epoch": 27.032295606633692, + "grad_norm": 0.1330031305551529, + "learning_rate": 3.9896529352244715e-06, + "loss": 0.0137, + "step": 92910 + }, + { + "epoch": 27.035205120744834, + "grad_norm": 0.08556458353996277, + "learning_rate": 3.986846073845243e-06, + "loss": 0.0129, + "step": 92920 + }, + { + "epoch": 27.03811463485598, + "grad_norm": 0.17518799006938934, + "learning_rate": 3.984040048842372e-06, + "loss": 0.0127, + "step": 92930 + }, + { + "epoch": 27.04102414896712, + "grad_norm": 0.12118197232484818, + "learning_rate": 3.981234860428963e-06, + "loss": 0.0104, + "step": 92940 + }, + { + "epoch": 27.043933663078267, + "grad_norm": 0.11654990911483765, + "learning_rate": 3.978430508818044e-06, + "loss": 0.0122, + "step": 92950 + }, + { + "epoch": 27.04684317718941, + "grad_norm": 0.09441301971673965, + "learning_rate": 3.9756269942225985e-06, + "loss": 0.0124, + "step": 92960 + }, + { + "epoch": 27.049752691300554, + "grad_norm": 0.12860102951526642, + "learning_rate": 3.972824316855524e-06, + "loss": 0.0099, + "step": 92970 + }, + { + "epoch": 27.052662205411696, + "grad_norm": 0.09998370707035065, + "learning_rate": 3.970022476929667e-06, + "loss": 0.0113, + "step": 92980 + }, + { + "epoch": 27.05557171952284, + "grad_norm": 0.13079224526882172, + "learning_rate": 3.9672214746578165e-06, + "loss": 0.0133, + "step": 92990 + }, + { + "epoch": 27.058481233633984, + "grad_norm": 0.07109879702329636, + "learning_rate": 3.964421310252682e-06, + "loss": 0.0125, + "step": 93000 + }, + { + "epoch": 27.061390747745126, + "grad_norm": 0.09731003642082214, + "learning_rate": 3.961621983926923e-06, + "loss": 0.0116, + "step": 93010 + }, + { + "epoch": 27.06430026185627, + "grad_norm": 0.09186958521604538, + "learning_rate": 3.958823495893124e-06, + "loss": 0.0116, + "step": 93020 + }, + { + "epoch": 27.067209775967413, + "grad_norm": 0.15172959864139557, + "learning_rate": 3.95602584636382e-06, + "loss": 0.0108, + "step": 93030 + }, + { + "epoch": 27.07011929007856, + "grad_norm": 0.09289747476577759, + "learning_rate": 3.953229035551464e-06, + "loss": 0.0117, + "step": 93040 + }, + { + "epoch": 27.0730288041897, + "grad_norm": 0.07643039524555206, + "learning_rate": 3.95043306366846e-06, + "loss": 0.0109, + "step": 93050 + }, + { + "epoch": 27.075938318300842, + "grad_norm": 0.10168536752462387, + "learning_rate": 3.94763793092715e-06, + "loss": 0.0125, + "step": 93060 + }, + { + "epoch": 27.078847832411988, + "grad_norm": 0.09128133952617645, + "learning_rate": 3.944843637539797e-06, + "loss": 0.0116, + "step": 93070 + }, + { + "epoch": 27.08175734652313, + "grad_norm": 0.08423581719398499, + "learning_rate": 3.942050183718606e-06, + "loss": 0.0113, + "step": 93080 + }, + { + "epoch": 27.084666860634275, + "grad_norm": 0.09448714554309845, + "learning_rate": 3.939257569675735e-06, + "loss": 0.0107, + "step": 93090 + }, + { + "epoch": 27.087576374745417, + "grad_norm": 0.13805058598518372, + "learning_rate": 3.936465795623257e-06, + "loss": 0.0112, + "step": 93100 + }, + { + "epoch": 27.090485888856563, + "grad_norm": 0.11928583681583405, + "learning_rate": 3.933674861773185e-06, + "loss": 0.0118, + "step": 93110 + }, + { + "epoch": 27.093395402967705, + "grad_norm": 0.12881624698638916, + "learning_rate": 3.930884768337476e-06, + "loss": 0.0134, + "step": 93120 + }, + { + "epoch": 27.096304917078847, + "grad_norm": 0.09364187717437744, + "learning_rate": 3.928095515528022e-06, + "loss": 0.0107, + "step": 93130 + }, + { + "epoch": 27.099214431189992, + "grad_norm": 0.34514328837394714, + "learning_rate": 3.9253071035566476e-06, + "loss": 0.0106, + "step": 93140 + }, + { + "epoch": 27.102123945301134, + "grad_norm": 0.08505402505397797, + "learning_rate": 3.9225195326351056e-06, + "loss": 0.0109, + "step": 93150 + }, + { + "epoch": 27.10503345941228, + "grad_norm": 0.13663069903850555, + "learning_rate": 3.919732802975102e-06, + "loss": 0.012, + "step": 93160 + }, + { + "epoch": 27.10794297352342, + "grad_norm": 0.12823836505413055, + "learning_rate": 3.916946914788272e-06, + "loss": 0.012, + "step": 93170 + }, + { + "epoch": 27.110852487634563, + "grad_norm": 0.127447247505188, + "learning_rate": 3.914161868286177e-06, + "loss": 0.01, + "step": 93180 + }, + { + "epoch": 27.11376200174571, + "grad_norm": 0.12690603733062744, + "learning_rate": 3.9113776636803296e-06, + "loss": 0.0112, + "step": 93190 + }, + { + "epoch": 27.11667151585685, + "grad_norm": 0.09973982721567154, + "learning_rate": 3.908594301182172e-06, + "loss": 0.0109, + "step": 93200 + }, + { + "epoch": 27.119581029967996, + "grad_norm": 0.1418268382549286, + "learning_rate": 3.905811781003083e-06, + "loss": 0.0107, + "step": 93210 + }, + { + "epoch": 27.122490544079138, + "grad_norm": 0.12688018381595612, + "learning_rate": 3.9030301033543674e-06, + "loss": 0.0102, + "step": 93220 + }, + { + "epoch": 27.125400058190284, + "grad_norm": 0.1264631599187851, + "learning_rate": 3.900249268447283e-06, + "loss": 0.0106, + "step": 93230 + }, + { + "epoch": 27.128309572301426, + "grad_norm": 0.10690052062273026, + "learning_rate": 3.897469276493018e-06, + "loss": 0.0109, + "step": 93240 + }, + { + "epoch": 27.131219086412568, + "grad_norm": 0.12037681043148041, + "learning_rate": 3.894690127702689e-06, + "loss": 0.0113, + "step": 93250 + }, + { + "epoch": 27.134128600523713, + "grad_norm": 0.11801845580339432, + "learning_rate": 3.891911822287359e-06, + "loss": 0.0102, + "step": 93260 + }, + { + "epoch": 27.137038114634855, + "grad_norm": 0.10939665138721466, + "learning_rate": 3.889134360458014e-06, + "loss": 0.0096, + "step": 93270 + }, + { + "epoch": 27.139947628746, + "grad_norm": 0.08864489197731018, + "learning_rate": 3.88635774242559e-06, + "loss": 0.0139, + "step": 93280 + }, + { + "epoch": 27.142857142857142, + "grad_norm": 0.07263367623090744, + "learning_rate": 3.883581968400957e-06, + "loss": 0.0111, + "step": 93290 + }, + { + "epoch": 27.145766656968288, + "grad_norm": 0.15792590379714966, + "learning_rate": 3.8808070385949065e-06, + "loss": 0.0102, + "step": 93300 + }, + { + "epoch": 27.14867617107943, + "grad_norm": 0.09783598780632019, + "learning_rate": 3.878032953218183e-06, + "loss": 0.0112, + "step": 93310 + }, + { + "epoch": 27.15158568519057, + "grad_norm": 0.0786840096116066, + "learning_rate": 3.875259712481457e-06, + "loss": 0.0143, + "step": 93320 + }, + { + "epoch": 27.154495199301717, + "grad_norm": 0.13996651768684387, + "learning_rate": 3.872487316595342e-06, + "loss": 0.0123, + "step": 93330 + }, + { + "epoch": 27.15740471341286, + "grad_norm": 0.15401732921600342, + "learning_rate": 3.8697157657703755e-06, + "loss": 0.0137, + "step": 93340 + }, + { + "epoch": 27.160314227524005, + "grad_norm": 0.10242515057325363, + "learning_rate": 3.866945060217044e-06, + "loss": 0.0117, + "step": 93350 + }, + { + "epoch": 27.163223741635147, + "grad_norm": 0.11108355969190598, + "learning_rate": 3.864175200145767e-06, + "loss": 0.0144, + "step": 93360 + }, + { + "epoch": 27.166133255746292, + "grad_norm": 0.08687623590230942, + "learning_rate": 3.861406185766894e-06, + "loss": 0.0095, + "step": 93370 + }, + { + "epoch": 27.169042769857434, + "grad_norm": 0.07990217208862305, + "learning_rate": 3.858638017290704e-06, + "loss": 0.0102, + "step": 93380 + }, + { + "epoch": 27.171952283968576, + "grad_norm": 0.10726264119148254, + "learning_rate": 3.855870694927439e-06, + "loss": 0.0094, + "step": 93390 + }, + { + "epoch": 27.17486179807972, + "grad_norm": 0.09072327613830566, + "learning_rate": 3.853104218887249e-06, + "loss": 0.0114, + "step": 93400 + }, + { + "epoch": 27.177771312190863, + "grad_norm": 0.09515589475631714, + "learning_rate": 3.8503385893802275e-06, + "loss": 0.0111, + "step": 93410 + }, + { + "epoch": 27.18068082630201, + "grad_norm": 0.10817790031433105, + "learning_rate": 3.847573806616408e-06, + "loss": 0.0115, + "step": 93420 + }, + { + "epoch": 27.18359034041315, + "grad_norm": 0.13244600594043732, + "learning_rate": 3.8448098708057626e-06, + "loss": 0.0156, + "step": 93430 + }, + { + "epoch": 27.186499854524296, + "grad_norm": 0.12518276274204254, + "learning_rate": 3.842046782158189e-06, + "loss": 0.0108, + "step": 93440 + }, + { + "epoch": 27.189409368635438, + "grad_norm": 0.12219314277172089, + "learning_rate": 3.839284540883522e-06, + "loss": 0.0127, + "step": 93450 + }, + { + "epoch": 27.19231888274658, + "grad_norm": 0.15441487729549408, + "learning_rate": 3.836523147191541e-06, + "loss": 0.0113, + "step": 93460 + }, + { + "epoch": 27.195228396857726, + "grad_norm": 0.06515106558799744, + "learning_rate": 3.833762601291958e-06, + "loss": 0.0116, + "step": 93470 + }, + { + "epoch": 27.198137910968867, + "grad_norm": 0.0819002017378807, + "learning_rate": 3.831002903394414e-06, + "loss": 0.0119, + "step": 93480 + }, + { + "epoch": 27.201047425080013, + "grad_norm": 0.11505831032991409, + "learning_rate": 3.828244053708484e-06, + "loss": 0.012, + "step": 93490 + }, + { + "epoch": 27.203956939191155, + "grad_norm": 0.11051633954048157, + "learning_rate": 3.825486052443697e-06, + "loss": 0.0095, + "step": 93500 + }, + { + "epoch": 27.206866453302297, + "grad_norm": 0.09388643503189087, + "learning_rate": 3.822728899809501e-06, + "loss": 0.0132, + "step": 93510 + }, + { + "epoch": 27.209775967413442, + "grad_norm": 0.10651973634958267, + "learning_rate": 3.819972596015276e-06, + "loss": 0.0093, + "step": 93520 + }, + { + "epoch": 27.212685481524584, + "grad_norm": 0.08555954694747925, + "learning_rate": 3.817217141270351e-06, + "loss": 0.01, + "step": 93530 + }, + { + "epoch": 27.21559499563573, + "grad_norm": 0.11285668611526489, + "learning_rate": 3.8144625357839885e-06, + "loss": 0.0112, + "step": 93540 + }, + { + "epoch": 27.21850450974687, + "grad_norm": 0.09565701335668564, + "learning_rate": 3.8117087797653727e-06, + "loss": 0.0123, + "step": 93550 + }, + { + "epoch": 27.221414023858017, + "grad_norm": 0.10066480934619904, + "learning_rate": 3.808955873423646e-06, + "loss": 0.0128, + "step": 93560 + }, + { + "epoch": 27.22432353796916, + "grad_norm": 0.17633379995822906, + "learning_rate": 3.8062038169678588e-06, + "loss": 0.0127, + "step": 93570 + }, + { + "epoch": 27.2272330520803, + "grad_norm": 0.09696795791387558, + "learning_rate": 3.8034526106070244e-06, + "loss": 0.0111, + "step": 93580 + }, + { + "epoch": 27.230142566191446, + "grad_norm": 0.13921082019805908, + "learning_rate": 3.8007022545500706e-06, + "loss": 0.0109, + "step": 93590 + }, + { + "epoch": 27.23305208030259, + "grad_norm": 0.10837170481681824, + "learning_rate": 3.7979527490058707e-06, + "loss": 0.0111, + "step": 93600 + }, + { + "epoch": 27.235961594413734, + "grad_norm": 0.14415352046489716, + "learning_rate": 3.7952040941832356e-06, + "loss": 0.0087, + "step": 93610 + }, + { + "epoch": 27.238871108524876, + "grad_norm": 0.13389325141906738, + "learning_rate": 3.7924562902909018e-06, + "loss": 0.0125, + "step": 93620 + }, + { + "epoch": 27.24178062263602, + "grad_norm": 0.09202083200216293, + "learning_rate": 3.789709337537552e-06, + "loss": 0.0116, + "step": 93630 + }, + { + "epoch": 27.244690136747163, + "grad_norm": 0.07119566202163696, + "learning_rate": 3.7869632361317928e-06, + "loss": 0.012, + "step": 93640 + }, + { + "epoch": 27.247599650858305, + "grad_norm": 0.07909061014652252, + "learning_rate": 3.784217986282178e-06, + "loss": 0.0104, + "step": 93650 + }, + { + "epoch": 27.25050916496945, + "grad_norm": 0.11125337332487106, + "learning_rate": 3.781473588197192e-06, + "loss": 0.01, + "step": 93660 + }, + { + "epoch": 27.253418679080593, + "grad_norm": 0.11387079209089279, + "learning_rate": 3.778730042085251e-06, + "loss": 0.0115, + "step": 93670 + }, + { + "epoch": 27.256328193191738, + "grad_norm": 0.08182413130998611, + "learning_rate": 3.7759873481547057e-06, + "loss": 0.0123, + "step": 93680 + }, + { + "epoch": 27.25923770730288, + "grad_norm": 0.08462554961442947, + "learning_rate": 3.7732455066138493e-06, + "loss": 0.0132, + "step": 93690 + }, + { + "epoch": 27.262147221414025, + "grad_norm": 0.1540762186050415, + "learning_rate": 3.77050451767091e-06, + "loss": 0.0121, + "step": 93700 + }, + { + "epoch": 27.265056735525167, + "grad_norm": 0.1305098682641983, + "learning_rate": 3.7677643815340403e-06, + "loss": 0.0106, + "step": 93710 + }, + { + "epoch": 27.26796624963631, + "grad_norm": 0.08059406280517578, + "learning_rate": 3.7650250984113393e-06, + "loss": 0.0114, + "step": 93720 + }, + { + "epoch": 27.270875763747455, + "grad_norm": 0.12834946811199188, + "learning_rate": 3.7622866685108433e-06, + "loss": 0.0111, + "step": 93730 + }, + { + "epoch": 27.273785277858597, + "grad_norm": 0.14321938157081604, + "learning_rate": 3.759549092040512e-06, + "loss": 0.012, + "step": 93740 + }, + { + "epoch": 27.276694791969742, + "grad_norm": 0.11801843345165253, + "learning_rate": 3.7568123692082423e-06, + "loss": 0.0114, + "step": 93750 + }, + { + "epoch": 27.279604306080884, + "grad_norm": 0.1255379319190979, + "learning_rate": 3.7540765002218736e-06, + "loss": 0.0124, + "step": 93760 + }, + { + "epoch": 27.28251382019203, + "grad_norm": 0.1803712099790573, + "learning_rate": 3.7513414852891826e-06, + "loss": 0.011, + "step": 93770 + }, + { + "epoch": 27.28542333430317, + "grad_norm": 0.0934293270111084, + "learning_rate": 3.7486073246178714e-06, + "loss": 0.0109, + "step": 93780 + }, + { + "epoch": 27.288332848414314, + "grad_norm": 0.09717763215303421, + "learning_rate": 3.7458740184155726e-06, + "loss": 0.011, + "step": 93790 + }, + { + "epoch": 27.29124236252546, + "grad_norm": 0.10298573225736618, + "learning_rate": 3.74314156688988e-06, + "loss": 0.011, + "step": 93800 + }, + { + "epoch": 27.2941518766366, + "grad_norm": 0.06576182693243027, + "learning_rate": 3.7404099702482973e-06, + "loss": 0.009, + "step": 93810 + }, + { + "epoch": 27.297061390747746, + "grad_norm": 0.10978743433952332, + "learning_rate": 3.737679228698266e-06, + "loss": 0.0112, + "step": 93820 + }, + { + "epoch": 27.29997090485889, + "grad_norm": 0.0954269990324974, + "learning_rate": 3.7349493424471708e-06, + "loss": 0.013, + "step": 93830 + }, + { + "epoch": 27.30288041897003, + "grad_norm": 0.13360445201396942, + "learning_rate": 3.732220311702337e-06, + "loss": 0.0113, + "step": 93840 + }, + { + "epoch": 27.305789933081176, + "grad_norm": 0.07584881037473679, + "learning_rate": 3.729492136671008e-06, + "loss": 0.0116, + "step": 93850 + }, + { + "epoch": 27.308699447192318, + "grad_norm": 0.1658412218093872, + "learning_rate": 3.726764817560368e-06, + "loss": 0.0108, + "step": 93860 + }, + { + "epoch": 27.311608961303463, + "grad_norm": 0.0883476734161377, + "learning_rate": 3.7240383545775437e-06, + "loss": 0.0098, + "step": 93870 + }, + { + "epoch": 27.314518475414605, + "grad_norm": 0.09576579183340073, + "learning_rate": 3.7213127479295954e-06, + "loss": 0.0112, + "step": 93880 + }, + { + "epoch": 27.31742798952575, + "grad_norm": 0.09901919215917587, + "learning_rate": 3.7185879978235064e-06, + "loss": 0.0118, + "step": 93890 + }, + { + "epoch": 27.320337503636893, + "grad_norm": 0.07786305993795395, + "learning_rate": 3.7158641044662096e-06, + "loss": 0.0114, + "step": 93900 + }, + { + "epoch": 27.323247017748034, + "grad_norm": 0.09601613879203796, + "learning_rate": 3.713141068064569e-06, + "loss": 0.013, + "step": 93910 + }, + { + "epoch": 27.32615653185918, + "grad_norm": 0.10581165552139282, + "learning_rate": 3.7104188888253726e-06, + "loss": 0.0107, + "step": 93920 + }, + { + "epoch": 27.329066045970322, + "grad_norm": 0.14208471775054932, + "learning_rate": 3.7076975669553604e-06, + "loss": 0.0121, + "step": 93930 + }, + { + "epoch": 27.331975560081467, + "grad_norm": 0.1389925628900528, + "learning_rate": 3.7049771026611916e-06, + "loss": 0.0127, + "step": 93940 + }, + { + "epoch": 27.33488507419261, + "grad_norm": 0.1001415029168129, + "learning_rate": 3.7022574961494764e-06, + "loss": 0.0106, + "step": 93950 + }, + { + "epoch": 27.337794588303755, + "grad_norm": 0.12371004372835159, + "learning_rate": 3.6995387476267412e-06, + "loss": 0.0101, + "step": 93960 + }, + { + "epoch": 27.340704102414897, + "grad_norm": 0.1526404768228531, + "learning_rate": 3.6968208572994656e-06, + "loss": 0.0112, + "step": 93970 + }, + { + "epoch": 27.34361361652604, + "grad_norm": 0.05377037078142166, + "learning_rate": 3.694103825374046e-06, + "loss": 0.0109, + "step": 93980 + }, + { + "epoch": 27.346523130637184, + "grad_norm": 0.08587592095136642, + "learning_rate": 3.6913876520568296e-06, + "loss": 0.0108, + "step": 93990 + }, + { + "epoch": 27.349432644748326, + "grad_norm": 0.11606904864311218, + "learning_rate": 3.6886723375540958e-06, + "loss": 0.0115, + "step": 94000 + }, + { + "epoch": 27.35234215885947, + "grad_norm": 0.09281585365533829, + "learning_rate": 3.6859578820720434e-06, + "loss": 0.0105, + "step": 94010 + }, + { + "epoch": 27.355251672970613, + "grad_norm": 0.10341320186853409, + "learning_rate": 3.683244285816826e-06, + "loss": 0.011, + "step": 94020 + }, + { + "epoch": 27.35816118708176, + "grad_norm": 0.08330871164798737, + "learning_rate": 3.6805315489945245e-06, + "loss": 0.0096, + "step": 94030 + }, + { + "epoch": 27.3610707011929, + "grad_norm": 0.09992986917495728, + "learning_rate": 3.6778196718111485e-06, + "loss": 0.0106, + "step": 94040 + }, + { + "epoch": 27.363980215304043, + "grad_norm": 0.10565102845430374, + "learning_rate": 3.6751086544726457e-06, + "loss": 0.0155, + "step": 94050 + }, + { + "epoch": 27.36688972941519, + "grad_norm": 0.14161178469657898, + "learning_rate": 3.6723984971849024e-06, + "loss": 0.0126, + "step": 94060 + }, + { + "epoch": 27.36979924352633, + "grad_norm": 0.09731929749250412, + "learning_rate": 3.669689200153741e-06, + "loss": 0.0126, + "step": 94070 + }, + { + "epoch": 27.372708757637476, + "grad_norm": 0.09061943739652634, + "learning_rate": 3.666980763584912e-06, + "loss": 0.0123, + "step": 94080 + }, + { + "epoch": 27.375618271748618, + "grad_norm": 0.10220864415168762, + "learning_rate": 3.664273187684094e-06, + "loss": 0.0107, + "step": 94090 + }, + { + "epoch": 27.378527785859763, + "grad_norm": 0.07552880793809891, + "learning_rate": 3.661566472656927e-06, + "loss": 0.0134, + "step": 94100 + }, + { + "epoch": 27.381437299970905, + "grad_norm": 0.09306006133556366, + "learning_rate": 3.6588606187089597e-06, + "loss": 0.0108, + "step": 94110 + }, + { + "epoch": 27.384346814082047, + "grad_norm": 0.10749334841966629, + "learning_rate": 3.6561556260456775e-06, + "loss": 0.0098, + "step": 94120 + }, + { + "epoch": 27.387256328193192, + "grad_norm": 0.11530201882123947, + "learning_rate": 3.6534514948725145e-06, + "loss": 0.0117, + "step": 94130 + }, + { + "epoch": 27.390165842304334, + "grad_norm": 0.15295293927192688, + "learning_rate": 3.6507482253948334e-06, + "loss": 0.0107, + "step": 94140 + }, + { + "epoch": 27.39307535641548, + "grad_norm": 0.10071603953838348, + "learning_rate": 3.6480458178179244e-06, + "loss": 0.0115, + "step": 94150 + }, + { + "epoch": 27.395984870526622, + "grad_norm": 0.13424327969551086, + "learning_rate": 3.6453442723470173e-06, + "loss": 0.0102, + "step": 94160 + }, + { + "epoch": 27.398894384637764, + "grad_norm": 0.09991291165351868, + "learning_rate": 3.6426435891872774e-06, + "loss": 0.0103, + "step": 94170 + }, + { + "epoch": 27.40180389874891, + "grad_norm": 0.08748897165060043, + "learning_rate": 3.639943768543809e-06, + "loss": 0.0121, + "step": 94180 + }, + { + "epoch": 27.40471341286005, + "grad_norm": 0.11095276474952698, + "learning_rate": 3.6372448106216373e-06, + "loss": 0.0107, + "step": 94190 + }, + { + "epoch": 27.407622926971197, + "grad_norm": 0.1247183158993721, + "learning_rate": 3.634546715625734e-06, + "loss": 0.0119, + "step": 94200 + }, + { + "epoch": 27.41053244108234, + "grad_norm": 0.11863432824611664, + "learning_rate": 3.631849483761007e-06, + "loss": 0.0122, + "step": 94210 + }, + { + "epoch": 27.413441955193484, + "grad_norm": 0.07958199828863144, + "learning_rate": 3.629153115232285e-06, + "loss": 0.0092, + "step": 94220 + }, + { + "epoch": 27.416351469304626, + "grad_norm": 0.08630666136741638, + "learning_rate": 3.6264576102443462e-06, + "loss": 0.011, + "step": 94230 + }, + { + "epoch": 27.419260983415768, + "grad_norm": 0.11470384895801544, + "learning_rate": 3.62376296900189e-06, + "loss": 0.0107, + "step": 94240 + }, + { + "epoch": 27.422170497526913, + "grad_norm": 0.06836704164743423, + "learning_rate": 3.6210691917095613e-06, + "loss": 0.0125, + "step": 94250 + }, + { + "epoch": 27.425080011638055, + "grad_norm": 0.1486886739730835, + "learning_rate": 3.6183762785719303e-06, + "loss": 0.0115, + "step": 94260 + }, + { + "epoch": 27.4279895257492, + "grad_norm": 0.10942312330007553, + "learning_rate": 3.615684229793514e-06, + "loss": 0.012, + "step": 94270 + }, + { + "epoch": 27.430899039860343, + "grad_norm": 0.1345135122537613, + "learning_rate": 3.6129930455787446e-06, + "loss": 0.0123, + "step": 94280 + }, + { + "epoch": 27.433808553971488, + "grad_norm": 0.11267916858196259, + "learning_rate": 3.6103027261320075e-06, + "loss": 0.012, + "step": 94290 + }, + { + "epoch": 27.43671806808263, + "grad_norm": 0.12139949947595596, + "learning_rate": 3.607613271657616e-06, + "loss": 0.011, + "step": 94300 + }, + { + "epoch": 27.439627582193772, + "grad_norm": 0.1378583014011383, + "learning_rate": 3.60492468235981e-06, + "loss": 0.0116, + "step": 94310 + }, + { + "epoch": 27.442537096304918, + "grad_norm": 0.13149873912334442, + "learning_rate": 3.6022369584427755e-06, + "loss": 0.013, + "step": 94320 + }, + { + "epoch": 27.44544661041606, + "grad_norm": 0.0947219654917717, + "learning_rate": 3.5995501001106234e-06, + "loss": 0.0111, + "step": 94330 + }, + { + "epoch": 27.448356124527205, + "grad_norm": 0.08003505319356918, + "learning_rate": 3.5968641075674068e-06, + "loss": 0.0103, + "step": 94340 + }, + { + "epoch": 27.451265638638347, + "grad_norm": 0.10896062850952148, + "learning_rate": 3.5941789810171043e-06, + "loss": 0.0102, + "step": 94350 + }, + { + "epoch": 27.454175152749492, + "grad_norm": 0.06824241578578949, + "learning_rate": 3.591494720663636e-06, + "loss": 0.0102, + "step": 94360 + }, + { + "epoch": 27.457084666860634, + "grad_norm": 0.08297097682952881, + "learning_rate": 3.588811326710859e-06, + "loss": 0.0104, + "step": 94370 + }, + { + "epoch": 27.459994180971776, + "grad_norm": 0.12354122847318649, + "learning_rate": 3.5861287993625536e-06, + "loss": 0.0112, + "step": 94380 + }, + { + "epoch": 27.46290369508292, + "grad_norm": 0.18749909102916718, + "learning_rate": 3.5834471388224334e-06, + "loss": 0.0122, + "step": 94390 + }, + { + "epoch": 27.465813209194064, + "grad_norm": 0.11915509402751923, + "learning_rate": 3.5807663452941685e-06, + "loss": 0.0114, + "step": 94400 + }, + { + "epoch": 27.46872272330521, + "grad_norm": 0.11411920189857483, + "learning_rate": 3.5780864189813382e-06, + "loss": 0.0092, + "step": 94410 + }, + { + "epoch": 27.47163223741635, + "grad_norm": 0.3275917172431946, + "learning_rate": 3.575407360087463e-06, + "loss": 0.0129, + "step": 94420 + }, + { + "epoch": 27.474541751527497, + "grad_norm": 0.08812204003334045, + "learning_rate": 3.572729168816002e-06, + "loss": 0.0108, + "step": 94430 + }, + { + "epoch": 27.47745126563864, + "grad_norm": 0.10042457282543182, + "learning_rate": 3.5700518453703533e-06, + "loss": 0.0128, + "step": 94440 + }, + { + "epoch": 27.48036077974978, + "grad_norm": 0.1212499663233757, + "learning_rate": 3.5673753899538352e-06, + "loss": 0.0116, + "step": 94450 + }, + { + "epoch": 27.483270293860926, + "grad_norm": 0.1081278845667839, + "learning_rate": 3.5646998027697012e-06, + "loss": 0.01, + "step": 94460 + }, + { + "epoch": 27.486179807972068, + "grad_norm": 0.09333217144012451, + "learning_rate": 3.562025084021153e-06, + "loss": 0.011, + "step": 94470 + }, + { + "epoch": 27.489089322083213, + "grad_norm": 0.11088661849498749, + "learning_rate": 3.55935123391132e-06, + "loss": 0.0126, + "step": 94480 + }, + { + "epoch": 27.491998836194355, + "grad_norm": 0.12610630691051483, + "learning_rate": 3.556678252643254e-06, + "loss": 0.0139, + "step": 94490 + }, + { + "epoch": 27.4949083503055, + "grad_norm": 0.137221559882164, + "learning_rate": 3.5540061404199543e-06, + "loss": 0.0117, + "step": 94500 + }, + { + "epoch": 27.497817864416643, + "grad_norm": 0.10144466906785965, + "learning_rate": 3.5513348974443565e-06, + "loss": 0.0108, + "step": 94510 + }, + { + "epoch": 27.500727378527785, + "grad_norm": 0.19792458415031433, + "learning_rate": 3.5486645239193206e-06, + "loss": 0.0109, + "step": 94520 + }, + { + "epoch": 27.50363689263893, + "grad_norm": 0.10946356505155563, + "learning_rate": 3.5459950200476356e-06, + "loss": 0.0106, + "step": 94530 + }, + { + "epoch": 27.506546406750072, + "grad_norm": 0.10513915866613388, + "learning_rate": 3.5433263860320393e-06, + "loss": 0.0112, + "step": 94540 + }, + { + "epoch": 27.509455920861217, + "grad_norm": 0.11364459246397018, + "learning_rate": 3.5406586220751995e-06, + "loss": 0.0105, + "step": 94550 + }, + { + "epoch": 27.51236543497236, + "grad_norm": 0.06953130662441254, + "learning_rate": 3.53799172837971e-06, + "loss": 0.01, + "step": 94560 + }, + { + "epoch": 27.5152749490835, + "grad_norm": 0.10619767010211945, + "learning_rate": 3.535325705148109e-06, + "loss": 0.0109, + "step": 94570 + }, + { + "epoch": 27.518184463194647, + "grad_norm": 0.11646179109811783, + "learning_rate": 3.5326605525828558e-06, + "loss": 0.0106, + "step": 94580 + }, + { + "epoch": 27.52109397730579, + "grad_norm": 0.08058073371648788, + "learning_rate": 3.5299962708863563e-06, + "loss": 0.0103, + "step": 94590 + }, + { + "epoch": 27.524003491416934, + "grad_norm": 0.09547862410545349, + "learning_rate": 3.5273328602609504e-06, + "loss": 0.0115, + "step": 94600 + }, + { + "epoch": 27.526913005528076, + "grad_norm": 0.09992432594299316, + "learning_rate": 3.524670320908895e-06, + "loss": 0.011, + "step": 94610 + }, + { + "epoch": 27.52982251963922, + "grad_norm": 0.12963585555553436, + "learning_rate": 3.5220086530324027e-06, + "loss": 0.0105, + "step": 94620 + }, + { + "epoch": 27.532732033750364, + "grad_norm": 0.1298833042383194, + "learning_rate": 3.5193478568336e-06, + "loss": 0.0125, + "step": 94630 + }, + { + "epoch": 27.535641547861506, + "grad_norm": 0.11170578002929688, + "learning_rate": 3.516687932514568e-06, + "loss": 0.0114, + "step": 94640 + }, + { + "epoch": 27.53855106197265, + "grad_norm": 0.1346655786037445, + "learning_rate": 3.5140288802772973e-06, + "loss": 0.0109, + "step": 94650 + }, + { + "epoch": 27.541460576083793, + "grad_norm": 0.08231014013290405, + "learning_rate": 3.5113707003237355e-06, + "loss": 0.0106, + "step": 94660 + }, + { + "epoch": 27.54437009019494, + "grad_norm": 0.10566341131925583, + "learning_rate": 3.5087133928557526e-06, + "loss": 0.0102, + "step": 94670 + }, + { + "epoch": 27.54727960430608, + "grad_norm": 0.12202590703964233, + "learning_rate": 3.506056958075152e-06, + "loss": 0.0108, + "step": 94680 + }, + { + "epoch": 27.550189118417226, + "grad_norm": 0.11222313344478607, + "learning_rate": 3.5034013961836687e-06, + "loss": 0.0138, + "step": 94690 + }, + { + "epoch": 27.553098632528368, + "grad_norm": 0.2648308277130127, + "learning_rate": 3.500746707382977e-06, + "loss": 0.0115, + "step": 94700 + }, + { + "epoch": 27.55600814663951, + "grad_norm": 0.11764099448919296, + "learning_rate": 3.49809289187469e-06, + "loss": 0.0107, + "step": 94710 + }, + { + "epoch": 27.558917660750655, + "grad_norm": 0.09712879359722137, + "learning_rate": 3.4954399498603373e-06, + "loss": 0.01, + "step": 94720 + }, + { + "epoch": 27.561827174861797, + "grad_norm": 0.08830895274877548, + "learning_rate": 3.4927878815413953e-06, + "loss": 0.0106, + "step": 94730 + }, + { + "epoch": 27.564736688972943, + "grad_norm": 0.10168620944023132, + "learning_rate": 3.4901366871192784e-06, + "loss": 0.0117, + "step": 94740 + }, + { + "epoch": 27.567646203084085, + "grad_norm": 0.09465060383081436, + "learning_rate": 3.4874863667953224e-06, + "loss": 0.0099, + "step": 94750 + }, + { + "epoch": 27.57055571719523, + "grad_norm": 0.13587167859077454, + "learning_rate": 3.4848369207707955e-06, + "loss": 0.0121, + "step": 94760 + }, + { + "epoch": 27.573465231306372, + "grad_norm": 0.1235727071762085, + "learning_rate": 3.482188349246912e-06, + "loss": 0.0126, + "step": 94770 + }, + { + "epoch": 27.576374745417514, + "grad_norm": 0.1250971257686615, + "learning_rate": 3.4795406524248162e-06, + "loss": 0.011, + "step": 94780 + }, + { + "epoch": 27.57928425952866, + "grad_norm": 0.11103999614715576, + "learning_rate": 3.476893830505579e-06, + "loss": 0.0111, + "step": 94790 + }, + { + "epoch": 27.5821937736398, + "grad_norm": 0.12277573347091675, + "learning_rate": 3.4742478836902013e-06, + "loss": 0.012, + "step": 94800 + }, + { + "epoch": 27.585103287750947, + "grad_norm": 0.1308079957962036, + "learning_rate": 3.4716028121796404e-06, + "loss": 0.0117, + "step": 94810 + }, + { + "epoch": 27.58801280186209, + "grad_norm": 0.08796503394842148, + "learning_rate": 3.468958616174768e-06, + "loss": 0.0098, + "step": 94820 + }, + { + "epoch": 27.59092231597323, + "grad_norm": 0.10759902000427246, + "learning_rate": 3.4663152958763838e-06, + "loss": 0.0094, + "step": 94830 + }, + { + "epoch": 27.593831830084376, + "grad_norm": 0.15412397682666779, + "learning_rate": 3.4636728514852383e-06, + "loss": 0.0099, + "step": 94840 + }, + { + "epoch": 27.596741344195518, + "grad_norm": 0.13122279942035675, + "learning_rate": 3.4610312832020106e-06, + "loss": 0.01, + "step": 94850 + }, + { + "epoch": 27.599650858306664, + "grad_norm": 0.09805316478013992, + "learning_rate": 3.458390591227302e-06, + "loss": 0.0102, + "step": 94860 + }, + { + "epoch": 27.602560372417805, + "grad_norm": 0.07100105285644531, + "learning_rate": 3.455750775761666e-06, + "loss": 0.0113, + "step": 94870 + }, + { + "epoch": 27.60546988652895, + "grad_norm": 0.10157051682472229, + "learning_rate": 3.453111837005566e-06, + "loss": 0.0126, + "step": 94880 + }, + { + "epoch": 27.608379400640093, + "grad_norm": 0.10816706717014313, + "learning_rate": 3.4504737751594257e-06, + "loss": 0.0106, + "step": 94890 + }, + { + "epoch": 27.611288914751235, + "grad_norm": 0.11556370556354523, + "learning_rate": 3.4478365904235764e-06, + "loss": 0.0116, + "step": 94900 + }, + { + "epoch": 27.61419842886238, + "grad_norm": 0.17917726933956146, + "learning_rate": 3.4452002829983016e-06, + "loss": 0.0129, + "step": 94910 + }, + { + "epoch": 27.617107942973522, + "grad_norm": 0.06500552594661713, + "learning_rate": 3.4425648530838124e-06, + "loss": 0.0092, + "step": 94920 + }, + { + "epoch": 27.620017457084668, + "grad_norm": 0.10232625156641006, + "learning_rate": 3.4399303008802486e-06, + "loss": 0.0114, + "step": 94930 + }, + { + "epoch": 27.62292697119581, + "grad_norm": 0.08868725597858429, + "learning_rate": 3.4372966265876905e-06, + "loss": 0.0106, + "step": 94940 + }, + { + "epoch": 27.625836485306955, + "grad_norm": 0.12720325589179993, + "learning_rate": 3.4346638304061428e-06, + "loss": 0.0117, + "step": 94950 + }, + { + "epoch": 27.628745999418097, + "grad_norm": 0.07546389847993851, + "learning_rate": 3.4320319125355544e-06, + "loss": 0.0104, + "step": 94960 + }, + { + "epoch": 27.63165551352924, + "grad_norm": 0.10670248419046402, + "learning_rate": 3.429400873175803e-06, + "loss": 0.0119, + "step": 94970 + }, + { + "epoch": 27.634565027640384, + "grad_norm": 0.11871220916509628, + "learning_rate": 3.426770712526696e-06, + "loss": 0.0115, + "step": 94980 + }, + { + "epoch": 27.637474541751526, + "grad_norm": 0.1486886739730835, + "learning_rate": 3.424141430787972e-06, + "loss": 0.0119, + "step": 94990 + }, + { + "epoch": 27.640384055862672, + "grad_norm": 0.10083924233913422, + "learning_rate": 3.4215130281593125e-06, + "loss": 0.0115, + "step": 95000 + }, + { + "epoch": 27.643293569973814, + "grad_norm": 0.10269910842180252, + "learning_rate": 3.4188855048403316e-06, + "loss": 0.0104, + "step": 95010 + }, + { + "epoch": 27.64620308408496, + "grad_norm": 0.11101341992616653, + "learning_rate": 3.4162588610305643e-06, + "loss": 0.0086, + "step": 95020 + }, + { + "epoch": 27.6491125981961, + "grad_norm": 0.07598558068275452, + "learning_rate": 3.41363309692949e-06, + "loss": 0.0131, + "step": 95030 + }, + { + "epoch": 27.652022112307243, + "grad_norm": 0.1205715462565422, + "learning_rate": 3.4110082127365233e-06, + "loss": 0.0126, + "step": 95040 + }, + { + "epoch": 27.65493162641839, + "grad_norm": 0.12604796886444092, + "learning_rate": 3.408384208651002e-06, + "loss": 0.0103, + "step": 95050 + }, + { + "epoch": 27.65784114052953, + "grad_norm": 0.11156399548053741, + "learning_rate": 3.4057610848721965e-06, + "loss": 0.0115, + "step": 95060 + }, + { + "epoch": 27.660750654640676, + "grad_norm": 0.08702186495065689, + "learning_rate": 3.4031388415993227e-06, + "loss": 0.0104, + "step": 95070 + }, + { + "epoch": 27.663660168751818, + "grad_norm": 0.06884746998548508, + "learning_rate": 3.400517479031524e-06, + "loss": 0.0123, + "step": 95080 + }, + { + "epoch": 27.666569682862963, + "grad_norm": 0.13732360303401947, + "learning_rate": 3.3978969973678754e-06, + "loss": 0.0111, + "step": 95090 + }, + { + "epoch": 27.669479196974105, + "grad_norm": 0.12410221248865128, + "learning_rate": 3.395277396807372e-06, + "loss": 0.0135, + "step": 95100 + }, + { + "epoch": 27.672388711085247, + "grad_norm": 0.11395318806171417, + "learning_rate": 3.392658677548975e-06, + "loss": 0.0118, + "step": 95110 + }, + { + "epoch": 27.675298225196393, + "grad_norm": 0.16077199578285217, + "learning_rate": 3.39004083979155e-06, + "loss": 0.01, + "step": 95120 + }, + { + "epoch": 27.678207739307535, + "grad_norm": 0.11561659723520279, + "learning_rate": 3.3874238837339e-06, + "loss": 0.0107, + "step": 95130 + }, + { + "epoch": 27.68111725341868, + "grad_norm": 0.09501642733812332, + "learning_rate": 3.384807809574769e-06, + "loss": 0.01, + "step": 95140 + }, + { + "epoch": 27.684026767529822, + "grad_norm": 0.11177954077720642, + "learning_rate": 3.3821926175128355e-06, + "loss": 0.0123, + "step": 95150 + }, + { + "epoch": 27.686936281640968, + "grad_norm": 0.07810449600219727, + "learning_rate": 3.3795783077467015e-06, + "loss": 0.0117, + "step": 95160 + }, + { + "epoch": 27.68984579575211, + "grad_norm": 0.1867227703332901, + "learning_rate": 3.376964880474903e-06, + "loss": 0.0121, + "step": 95170 + }, + { + "epoch": 27.69275530986325, + "grad_norm": 0.09648644924163818, + "learning_rate": 3.374352335895917e-06, + "loss": 0.0102, + "step": 95180 + }, + { + "epoch": 27.695664823974397, + "grad_norm": 0.11608356237411499, + "learning_rate": 3.371740674208153e-06, + "loss": 0.01, + "step": 95190 + }, + { + "epoch": 27.69857433808554, + "grad_norm": 0.09437508136034012, + "learning_rate": 3.36912989560994e-06, + "loss": 0.012, + "step": 95200 + }, + { + "epoch": 27.701483852196684, + "grad_norm": 0.12725184857845306, + "learning_rate": 3.366520000299554e-06, + "loss": 0.0109, + "step": 95210 + }, + { + "epoch": 27.704393366307826, + "grad_norm": 0.11633815616369247, + "learning_rate": 3.363910988475205e-06, + "loss": 0.0108, + "step": 95220 + }, + { + "epoch": 27.707302880418972, + "grad_norm": 0.06640446186065674, + "learning_rate": 3.3613028603350203e-06, + "loss": 0.0122, + "step": 95230 + }, + { + "epoch": 27.710212394530114, + "grad_norm": 0.11604632437229156, + "learning_rate": 3.358695616077078e-06, + "loss": 0.011, + "step": 95240 + }, + { + "epoch": 27.713121908641256, + "grad_norm": 0.12904050946235657, + "learning_rate": 3.356089255899375e-06, + "loss": 0.0124, + "step": 95250 + }, + { + "epoch": 27.7160314227524, + "grad_norm": 0.11814986914396286, + "learning_rate": 3.3534837799998524e-06, + "loss": 0.0103, + "step": 95260 + }, + { + "epoch": 27.718940936863543, + "grad_norm": 0.09813809394836426, + "learning_rate": 3.350879188576374e-06, + "loss": 0.0118, + "step": 95270 + }, + { + "epoch": 27.72185045097469, + "grad_norm": 0.10894063115119934, + "learning_rate": 3.348275481826746e-06, + "loss": 0.0116, + "step": 95280 + }, + { + "epoch": 27.72475996508583, + "grad_norm": 0.09469447284936905, + "learning_rate": 3.345672659948698e-06, + "loss": 0.0109, + "step": 95290 + }, + { + "epoch": 27.727669479196972, + "grad_norm": 0.14661753177642822, + "learning_rate": 3.3430707231398994e-06, + "loss": 0.0148, + "step": 95300 + }, + { + "epoch": 27.730578993308118, + "grad_norm": 0.1303730309009552, + "learning_rate": 3.340469671597954e-06, + "loss": 0.0106, + "step": 95310 + }, + { + "epoch": 27.73348850741926, + "grad_norm": 0.09840922057628632, + "learning_rate": 3.3378695055203866e-06, + "loss": 0.0108, + "step": 95320 + }, + { + "epoch": 27.736398021530405, + "grad_norm": 0.0980444848537445, + "learning_rate": 3.335270225104667e-06, + "loss": 0.0109, + "step": 95330 + }, + { + "epoch": 27.739307535641547, + "grad_norm": 0.13783204555511475, + "learning_rate": 3.332671830548198e-06, + "loss": 0.0118, + "step": 95340 + }, + { + "epoch": 27.742217049752693, + "grad_norm": 0.14507131278514862, + "learning_rate": 3.3300743220483037e-06, + "loss": 0.0124, + "step": 95350 + }, + { + "epoch": 27.745126563863835, + "grad_norm": 0.08336365967988968, + "learning_rate": 3.327477699802248e-06, + "loss": 0.0093, + "step": 95360 + }, + { + "epoch": 27.748036077974977, + "grad_norm": 0.13848307728767395, + "learning_rate": 3.3248819640072285e-06, + "loss": 0.0106, + "step": 95370 + }, + { + "epoch": 27.750945592086122, + "grad_norm": 0.16974091529846191, + "learning_rate": 3.322287114860379e-06, + "loss": 0.0133, + "step": 95380 + }, + { + "epoch": 27.753855106197264, + "grad_norm": 0.09915503114461899, + "learning_rate": 3.3196931525587566e-06, + "loss": 0.0101, + "step": 95390 + }, + { + "epoch": 27.75676462030841, + "grad_norm": 0.12499987334012985, + "learning_rate": 3.317100077299347e-06, + "loss": 0.0101, + "step": 95400 + }, + { + "epoch": 27.75967413441955, + "grad_norm": 0.08609102666378021, + "learning_rate": 3.3145078892790947e-06, + "loss": 0.0114, + "step": 95410 + }, + { + "epoch": 27.762583648530697, + "grad_norm": 0.14108934998512268, + "learning_rate": 3.311916588694852e-06, + "loss": 0.0103, + "step": 95420 + }, + { + "epoch": 27.76549316264184, + "grad_norm": 0.11490854620933533, + "learning_rate": 3.3093261757434035e-06, + "loss": 0.0099, + "step": 95430 + }, + { + "epoch": 27.76840267675298, + "grad_norm": 0.09430652856826782, + "learning_rate": 3.3067366506214814e-06, + "loss": 0.0107, + "step": 95440 + }, + { + "epoch": 27.771312190864126, + "grad_norm": 0.08889011293649673, + "learning_rate": 3.304148013525747e-06, + "loss": 0.0114, + "step": 95450 + }, + { + "epoch": 27.774221704975268, + "grad_norm": 0.09472872316837311, + "learning_rate": 3.3015602646527843e-06, + "loss": 0.0109, + "step": 95460 + }, + { + "epoch": 27.777131219086414, + "grad_norm": 0.06845200061798096, + "learning_rate": 3.2989734041991114e-06, + "loss": 0.0108, + "step": 95470 + }, + { + "epoch": 27.780040733197556, + "grad_norm": 0.14440305531024933, + "learning_rate": 3.2963874323611897e-06, + "loss": 0.0111, + "step": 95480 + }, + { + "epoch": 27.7829502473087, + "grad_norm": 0.13902699947357178, + "learning_rate": 3.2938023493354086e-06, + "loss": 0.0121, + "step": 95490 + }, + { + "epoch": 27.785859761419843, + "grad_norm": 0.06112741306424141, + "learning_rate": 3.291218155318081e-06, + "loss": 0.0097, + "step": 95500 + }, + { + "epoch": 27.788769275530985, + "grad_norm": 0.10871411859989166, + "learning_rate": 3.2886348505054633e-06, + "loss": 0.0114, + "step": 95510 + }, + { + "epoch": 27.79167878964213, + "grad_norm": 0.1257927268743515, + "learning_rate": 3.286052435093745e-06, + "loss": 0.0117, + "step": 95520 + }, + { + "epoch": 27.794588303753272, + "grad_norm": 0.15815994143486023, + "learning_rate": 3.283470909279038e-06, + "loss": 0.0129, + "step": 95530 + }, + { + "epoch": 27.797497817864418, + "grad_norm": 0.0984252318739891, + "learning_rate": 3.2808902732573896e-06, + "loss": 0.0131, + "step": 95540 + }, + { + "epoch": 27.80040733197556, + "grad_norm": 0.10689353942871094, + "learning_rate": 3.278310527224784e-06, + "loss": 0.0114, + "step": 95550 + }, + { + "epoch": 27.8033168460867, + "grad_norm": 0.05862057954072952, + "learning_rate": 3.2757316713771444e-06, + "loss": 0.0095, + "step": 95560 + }, + { + "epoch": 27.806226360197847, + "grad_norm": 0.1461976319551468, + "learning_rate": 3.2731537059103054e-06, + "loss": 0.0092, + "step": 95570 + }, + { + "epoch": 27.80913587430899, + "grad_norm": 0.10165470093488693, + "learning_rate": 3.2705766310200557e-06, + "loss": 0.0111, + "step": 95580 + }, + { + "epoch": 27.812045388420135, + "grad_norm": 0.20507851243019104, + "learning_rate": 3.2680004469021003e-06, + "loss": 0.0102, + "step": 95590 + }, + { + "epoch": 27.814954902531277, + "grad_norm": 0.08794418722391129, + "learning_rate": 3.265425153752086e-06, + "loss": 0.0114, + "step": 95600 + }, + { + "epoch": 27.817864416642422, + "grad_norm": 0.09608937054872513, + "learning_rate": 3.2628507517655927e-06, + "loss": 0.0115, + "step": 95610 + }, + { + "epoch": 27.820773930753564, + "grad_norm": 0.07978928089141846, + "learning_rate": 3.2602772411381243e-06, + "loss": 0.0132, + "step": 95620 + }, + { + "epoch": 27.823683444864706, + "grad_norm": 0.08332649618387222, + "learning_rate": 3.2577046220651264e-06, + "loss": 0.0113, + "step": 95630 + }, + { + "epoch": 27.82659295897585, + "grad_norm": 0.08877512067556381, + "learning_rate": 3.2551328947419675e-06, + "loss": 0.0114, + "step": 95640 + }, + { + "epoch": 27.829502473086993, + "grad_norm": 0.13399377465248108, + "learning_rate": 3.252562059363959e-06, + "loss": 0.0134, + "step": 95650 + }, + { + "epoch": 27.83241198719814, + "grad_norm": 0.14602626860141754, + "learning_rate": 3.249992116126331e-06, + "loss": 0.0126, + "step": 95660 + }, + { + "epoch": 27.83532150130928, + "grad_norm": 0.09860152006149292, + "learning_rate": 3.24742306522426e-06, + "loss": 0.0114, + "step": 95670 + }, + { + "epoch": 27.838231015420426, + "grad_norm": 0.07603809982538223, + "learning_rate": 3.2448549068528505e-06, + "loss": 0.0117, + "step": 95680 + }, + { + "epoch": 27.841140529531568, + "grad_norm": 0.14143702387809753, + "learning_rate": 3.2422876412071324e-06, + "loss": 0.0122, + "step": 95690 + }, + { + "epoch": 27.84405004364271, + "grad_norm": 0.08508665859699249, + "learning_rate": 3.2397212684820665e-06, + "loss": 0.0108, + "step": 95700 + }, + { + "epoch": 27.846959557753856, + "grad_norm": 0.3370261788368225, + "learning_rate": 3.237155788872569e-06, + "loss": 0.0098, + "step": 95710 + }, + { + "epoch": 27.849869071864997, + "grad_norm": 0.13603205978870392, + "learning_rate": 3.2345912025734586e-06, + "loss": 0.0112, + "step": 95720 + }, + { + "epoch": 27.852778585976143, + "grad_norm": 0.10424172878265381, + "learning_rate": 3.2320275097794997e-06, + "loss": 0.0118, + "step": 95730 + }, + { + "epoch": 27.855688100087285, + "grad_norm": 0.09976571053266525, + "learning_rate": 3.229464710685389e-06, + "loss": 0.0101, + "step": 95740 + }, + { + "epoch": 27.85859761419843, + "grad_norm": 0.14630518853664398, + "learning_rate": 3.2269028054857576e-06, + "loss": 0.0117, + "step": 95750 + }, + { + "epoch": 27.861507128309572, + "grad_norm": 0.08400361984968185, + "learning_rate": 3.2243417943751623e-06, + "loss": 0.011, + "step": 95760 + }, + { + "epoch": 27.864416642420714, + "grad_norm": 0.10217839479446411, + "learning_rate": 3.221781677548093e-06, + "loss": 0.0117, + "step": 95770 + }, + { + "epoch": 27.86732615653186, + "grad_norm": 0.1014319434762001, + "learning_rate": 3.2192224551989734e-06, + "loss": 0.011, + "step": 95780 + }, + { + "epoch": 27.870235670643, + "grad_norm": 0.10091187059879303, + "learning_rate": 3.2166641275221688e-06, + "loss": 0.012, + "step": 95790 + }, + { + "epoch": 27.873145184754147, + "grad_norm": 0.074745312333107, + "learning_rate": 3.2141066947119547e-06, + "loss": 0.0105, + "step": 95800 + }, + { + "epoch": 27.87605469886529, + "grad_norm": 0.11889063566923141, + "learning_rate": 3.211550156962556e-06, + "loss": 0.011, + "step": 95810 + }, + { + "epoch": 27.878964212976435, + "grad_norm": 0.12420188635587692, + "learning_rate": 3.208994514468131e-06, + "loss": 0.0095, + "step": 95820 + }, + { + "epoch": 27.881873727087576, + "grad_norm": 0.09920429438352585, + "learning_rate": 3.206439767422759e-06, + "loss": 0.0116, + "step": 95830 + }, + { + "epoch": 27.88478324119872, + "grad_norm": 0.1052580177783966, + "learning_rate": 3.203885916020452e-06, + "loss": 0.0122, + "step": 95840 + }, + { + "epoch": 27.887692755309864, + "grad_norm": 0.112399160861969, + "learning_rate": 3.2013329604551628e-06, + "loss": 0.0133, + "step": 95850 + }, + { + "epoch": 27.890602269421006, + "grad_norm": 0.0980461910367012, + "learning_rate": 3.198780900920773e-06, + "loss": 0.0132, + "step": 95860 + }, + { + "epoch": 27.89351178353215, + "grad_norm": 0.08061207085847855, + "learning_rate": 3.1962297376110896e-06, + "loss": 0.0125, + "step": 95870 + }, + { + "epoch": 27.896421297643293, + "grad_norm": 0.07072219997644424, + "learning_rate": 3.1936794707198635e-06, + "loss": 0.0095, + "step": 95880 + }, + { + "epoch": 27.89933081175444, + "grad_norm": 0.09454259276390076, + "learning_rate": 3.1911301004407622e-06, + "loss": 0.0108, + "step": 95890 + }, + { + "epoch": 27.90224032586558, + "grad_norm": 0.07414795458316803, + "learning_rate": 3.188581626967402e-06, + "loss": 0.0121, + "step": 95900 + }, + { + "epoch": 27.905149839976723, + "grad_norm": 0.11219244450330734, + "learning_rate": 3.1860340504933155e-06, + "loss": 0.0096, + "step": 95910 + }, + { + "epoch": 27.908059354087868, + "grad_norm": 0.1459604650735855, + "learning_rate": 3.1834873712119773e-06, + "loss": 0.0114, + "step": 95920 + }, + { + "epoch": 27.91096886819901, + "grad_norm": 0.09956956654787064, + "learning_rate": 3.1809415893167944e-06, + "loss": 0.0108, + "step": 95930 + }, + { + "epoch": 27.913878382310156, + "grad_norm": 0.12410473823547363, + "learning_rate": 3.1783967050010954e-06, + "loss": 0.012, + "step": 95940 + }, + { + "epoch": 27.916787896421297, + "grad_norm": 0.09649443626403809, + "learning_rate": 3.1758527184581544e-06, + "loss": 0.0121, + "step": 95950 + }, + { + "epoch": 27.91969741053244, + "grad_norm": 0.10246909409761429, + "learning_rate": 3.173309629881163e-06, + "loss": 0.0131, + "step": 95960 + }, + { + "epoch": 27.922606924643585, + "grad_norm": 0.10874198377132416, + "learning_rate": 3.170767439463256e-06, + "loss": 0.0099, + "step": 95970 + }, + { + "epoch": 27.925516438754727, + "grad_norm": 0.10506575554609299, + "learning_rate": 3.1682261473974995e-06, + "loss": 0.0118, + "step": 95980 + }, + { + "epoch": 27.928425952865872, + "grad_norm": 0.09434299916028976, + "learning_rate": 3.165685753876885e-06, + "loss": 0.0117, + "step": 95990 + }, + { + "epoch": 27.931335466977014, + "grad_norm": 0.15947142243385315, + "learning_rate": 3.1631462590943337e-06, + "loss": 0.011, + "step": 96000 + }, + { + "epoch": 27.93424498108816, + "grad_norm": 0.12598560750484467, + "learning_rate": 3.160607663242706e-06, + "loss": 0.0101, + "step": 96010 + }, + { + "epoch": 27.9371544951993, + "grad_norm": 0.0969325453042984, + "learning_rate": 3.1580699665147987e-06, + "loss": 0.0121, + "step": 96020 + }, + { + "epoch": 27.940064009310444, + "grad_norm": 0.09397216886281967, + "learning_rate": 3.155533169103323e-06, + "loss": 0.011, + "step": 96030 + }, + { + "epoch": 27.94297352342159, + "grad_norm": 0.12951764464378357, + "learning_rate": 3.1529972712009358e-06, + "loss": 0.0094, + "step": 96040 + }, + { + "epoch": 27.94588303753273, + "grad_norm": 0.10417152941226959, + "learning_rate": 3.1504622730002287e-06, + "loss": 0.0108, + "step": 96050 + }, + { + "epoch": 27.948792551643876, + "grad_norm": 0.09746059030294418, + "learning_rate": 3.147928174693709e-06, + "loss": 0.0113, + "step": 96060 + }, + { + "epoch": 27.95170206575502, + "grad_norm": 0.07501856982707977, + "learning_rate": 3.145394976473826e-06, + "loss": 0.0117, + "step": 96070 + }, + { + "epoch": 27.954611579866164, + "grad_norm": 0.16481079161167145, + "learning_rate": 3.1428626785329594e-06, + "loss": 0.0096, + "step": 96080 + }, + { + "epoch": 27.957521093977306, + "grad_norm": 0.22736619412899017, + "learning_rate": 3.1403312810634255e-06, + "loss": 0.0108, + "step": 96090 + }, + { + "epoch": 27.960430608088448, + "grad_norm": 0.10124709457159042, + "learning_rate": 3.137800784257465e-06, + "loss": 0.0111, + "step": 96100 + }, + { + "epoch": 27.963340122199593, + "grad_norm": 0.1162559986114502, + "learning_rate": 3.1352711883072434e-06, + "loss": 0.0099, + "step": 96110 + }, + { + "epoch": 27.966249636310735, + "grad_norm": 0.08285607397556305, + "learning_rate": 3.132742493404881e-06, + "loss": 0.0093, + "step": 96120 + }, + { + "epoch": 27.96915915042188, + "grad_norm": 0.23184438049793243, + "learning_rate": 3.1302146997424114e-06, + "loss": 0.0102, + "step": 96130 + }, + { + "epoch": 27.972068664533023, + "grad_norm": 0.15531349182128906, + "learning_rate": 3.127687807511796e-06, + "loss": 0.0092, + "step": 96140 + }, + { + "epoch": 27.974978178644168, + "grad_norm": 0.09367205202579498, + "learning_rate": 3.1251618169049413e-06, + "loss": 0.0091, + "step": 96150 + }, + { + "epoch": 27.97788769275531, + "grad_norm": 0.0954299196600914, + "learning_rate": 3.1226367281136835e-06, + "loss": 0.0105, + "step": 96160 + }, + { + "epoch": 27.980797206866452, + "grad_norm": 0.08079216629266739, + "learning_rate": 3.12011254132978e-06, + "loss": 0.0107, + "step": 96170 + }, + { + "epoch": 27.983706720977597, + "grad_norm": 0.13292697072029114, + "learning_rate": 3.1175892567449316e-06, + "loss": 0.0107, + "step": 96180 + }, + { + "epoch": 27.98661623508874, + "grad_norm": 0.11988982558250427, + "learning_rate": 3.1150668745507573e-06, + "loss": 0.0104, + "step": 96190 + }, + { + "epoch": 27.989525749199885, + "grad_norm": 0.0814858004450798, + "learning_rate": 3.112545394938825e-06, + "loss": 0.0106, + "step": 96200 + }, + { + "epoch": 27.992435263311027, + "grad_norm": 0.20008736848831177, + "learning_rate": 3.110024818100615e-06, + "loss": 0.0106, + "step": 96210 + }, + { + "epoch": 27.995344777422172, + "grad_norm": 0.10253681242465973, + "learning_rate": 3.107505144227553e-06, + "loss": 0.0119, + "step": 96220 + }, + { + "epoch": 27.998254291533314, + "grad_norm": 0.12773674726486206, + "learning_rate": 3.104986373510996e-06, + "loss": 0.0095, + "step": 96230 + }, + { + "epoch": 28.001163805644456, + "grad_norm": 0.15294595062732697, + "learning_rate": 3.1024685061422196e-06, + "loss": 0.0093, + "step": 96240 + }, + { + "epoch": 28.0040733197556, + "grad_norm": 0.1709751933813095, + "learning_rate": 3.0999515423124473e-06, + "loss": 0.0118, + "step": 96250 + }, + { + "epoch": 28.006982833866743, + "grad_norm": 0.18763971328735352, + "learning_rate": 3.0974354822128174e-06, + "loss": 0.011, + "step": 96260 + }, + { + "epoch": 28.00989234797789, + "grad_norm": 0.0964696928858757, + "learning_rate": 3.094920326034417e-06, + "loss": 0.011, + "step": 96270 + }, + { + "epoch": 28.01280186208903, + "grad_norm": 0.09700702130794525, + "learning_rate": 3.092406073968247e-06, + "loss": 0.0123, + "step": 96280 + }, + { + "epoch": 28.015711376200173, + "grad_norm": 0.08651600778102875, + "learning_rate": 3.089892726205257e-06, + "loss": 0.0101, + "step": 96290 + }, + { + "epoch": 28.01862089031132, + "grad_norm": 0.05795402079820633, + "learning_rate": 3.0873802829363108e-06, + "loss": 0.0107, + "step": 96300 + }, + { + "epoch": 28.02153040442246, + "grad_norm": 0.1959853172302246, + "learning_rate": 3.0848687443522145e-06, + "loss": 0.01, + "step": 96310 + }, + { + "epoch": 28.024439918533606, + "grad_norm": 0.10285027325153351, + "learning_rate": 3.0823581106437093e-06, + "loss": 0.0111, + "step": 96320 + }, + { + "epoch": 28.027349432644748, + "grad_norm": 0.10721846669912338, + "learning_rate": 3.079848382001451e-06, + "loss": 0.011, + "step": 96330 + }, + { + "epoch": 28.030258946755893, + "grad_norm": 0.09539555013179779, + "learning_rate": 3.0773395586160423e-06, + "loss": 0.0102, + "step": 96340 + }, + { + "epoch": 28.033168460867035, + "grad_norm": 0.09827952086925507, + "learning_rate": 3.074831640678014e-06, + "loss": 0.0088, + "step": 96350 + }, + { + "epoch": 28.036077974978177, + "grad_norm": 0.09395992755889893, + "learning_rate": 3.0723246283778256e-06, + "loss": 0.0102, + "step": 96360 + }, + { + "epoch": 28.038987489089322, + "grad_norm": 0.11089697480201721, + "learning_rate": 3.069818521905861e-06, + "loss": 0.0122, + "step": 96370 + }, + { + "epoch": 28.041897003200464, + "grad_norm": 0.06394951790571213, + "learning_rate": 3.0673133214524467e-06, + "loss": 0.0122, + "step": 96380 + }, + { + "epoch": 28.04480651731161, + "grad_norm": 0.10737595707178116, + "learning_rate": 3.0648090272078404e-06, + "loss": 0.0109, + "step": 96390 + }, + { + "epoch": 28.047716031422752, + "grad_norm": 0.09763885289430618, + "learning_rate": 3.062305639362225e-06, + "loss": 0.0091, + "step": 96400 + }, + { + "epoch": 28.050625545533897, + "grad_norm": 0.12139733880758286, + "learning_rate": 3.059803158105705e-06, + "loss": 0.0122, + "step": 96410 + }, + { + "epoch": 28.05353505964504, + "grad_norm": 0.14718031883239746, + "learning_rate": 3.0573015836283454e-06, + "loss": 0.0123, + "step": 96420 + }, + { + "epoch": 28.05644457375618, + "grad_norm": 0.13033831119537354, + "learning_rate": 3.054800916120115e-06, + "loss": 0.0115, + "step": 96430 + }, + { + "epoch": 28.059354087867327, + "grad_norm": 0.07716056704521179, + "learning_rate": 3.0523011557709195e-06, + "loss": 0.0121, + "step": 96440 + }, + { + "epoch": 28.06226360197847, + "grad_norm": 0.12140464782714844, + "learning_rate": 3.0498023027706036e-06, + "loss": 0.0123, + "step": 96450 + }, + { + "epoch": 28.065173116089614, + "grad_norm": 0.13308554887771606, + "learning_rate": 3.0473043573089414e-06, + "loss": 0.0118, + "step": 96460 + }, + { + "epoch": 28.068082630200756, + "grad_norm": 0.1049046739935875, + "learning_rate": 3.044807319575634e-06, + "loss": 0.0114, + "step": 96470 + }, + { + "epoch": 28.0709921443119, + "grad_norm": 0.0995122641324997, + "learning_rate": 3.0423111897603085e-06, + "loss": 0.0116, + "step": 96480 + }, + { + "epoch": 28.073901658423043, + "grad_norm": 0.08768175542354584, + "learning_rate": 3.039815968052534e-06, + "loss": 0.0093, + "step": 96490 + }, + { + "epoch": 28.076811172534185, + "grad_norm": 0.0760047435760498, + "learning_rate": 3.0373216546418105e-06, + "loss": 0.0101, + "step": 96500 + }, + { + "epoch": 28.07972068664533, + "grad_norm": 0.09915129840373993, + "learning_rate": 3.034828249717556e-06, + "loss": 0.0125, + "step": 96510 + }, + { + "epoch": 28.082630200756473, + "grad_norm": 0.1715427041053772, + "learning_rate": 3.032335753469135e-06, + "loss": 0.0094, + "step": 96520 + }, + { + "epoch": 28.08553971486762, + "grad_norm": 0.13833710551261902, + "learning_rate": 3.029844166085835e-06, + "loss": 0.0109, + "step": 96530 + }, + { + "epoch": 28.08844922897876, + "grad_norm": 0.07223451137542725, + "learning_rate": 3.027353487756872e-06, + "loss": 0.0115, + "step": 96540 + }, + { + "epoch": 28.091358743089906, + "grad_norm": 0.10387324541807175, + "learning_rate": 3.0248637186714027e-06, + "loss": 0.0117, + "step": 96550 + }, + { + "epoch": 28.094268257201048, + "grad_norm": 0.12927110493183136, + "learning_rate": 3.0223748590184997e-06, + "loss": 0.0108, + "step": 96560 + }, + { + "epoch": 28.09717777131219, + "grad_norm": 0.10737679898738861, + "learning_rate": 3.019886908987186e-06, + "loss": 0.0098, + "step": 96570 + }, + { + "epoch": 28.100087285423335, + "grad_norm": 0.09903975576162338, + "learning_rate": 3.0173998687663963e-06, + "loss": 0.012, + "step": 96580 + }, + { + "epoch": 28.102996799534477, + "grad_norm": 0.09906672686338425, + "learning_rate": 3.0149137385450104e-06, + "loss": 0.0101, + "step": 96590 + }, + { + "epoch": 28.105906313645622, + "grad_norm": 0.1351672261953354, + "learning_rate": 3.0124285185118295e-06, + "loss": 0.0115, + "step": 96600 + }, + { + "epoch": 28.108815827756764, + "grad_norm": 0.1094542145729065, + "learning_rate": 3.0099442088555906e-06, + "loss": 0.0113, + "step": 96610 + }, + { + "epoch": 28.11172534186791, + "grad_norm": 0.07726401090621948, + "learning_rate": 3.007460809764965e-06, + "loss": 0.0123, + "step": 96620 + }, + { + "epoch": 28.11463485597905, + "grad_norm": 0.14052297174930573, + "learning_rate": 3.004978321428544e-06, + "loss": 0.0115, + "step": 96630 + }, + { + "epoch": 28.117544370090194, + "grad_norm": 0.13830910623073578, + "learning_rate": 3.0024967440348585e-06, + "loss": 0.0145, + "step": 96640 + }, + { + "epoch": 28.12045388420134, + "grad_norm": 0.22034676373004913, + "learning_rate": 3.0000160777723728e-06, + "loss": 0.0096, + "step": 96650 + }, + { + "epoch": 28.12336339831248, + "grad_norm": 0.08394941687583923, + "learning_rate": 2.997536322829472e-06, + "loss": 0.0116, + "step": 96660 + }, + { + "epoch": 28.126272912423627, + "grad_norm": 0.07819589972496033, + "learning_rate": 2.995057479394476e-06, + "loss": 0.0088, + "step": 96670 + }, + { + "epoch": 28.12918242653477, + "grad_norm": 0.12791401147842407, + "learning_rate": 2.9925795476556378e-06, + "loss": 0.013, + "step": 96680 + }, + { + "epoch": 28.13209194064591, + "grad_norm": 0.16163764894008636, + "learning_rate": 2.9901025278011444e-06, + "loss": 0.0107, + "step": 96690 + }, + { + "epoch": 28.135001454757056, + "grad_norm": 0.1316872537136078, + "learning_rate": 2.987626420019107e-06, + "loss": 0.0107, + "step": 96700 + }, + { + "epoch": 28.137910968868198, + "grad_norm": 0.5938172340393066, + "learning_rate": 2.985151224497559e-06, + "loss": 0.0125, + "step": 96710 + }, + { + "epoch": 28.140820482979343, + "grad_norm": 0.18991714715957642, + "learning_rate": 2.982676941424492e-06, + "loss": 0.0127, + "step": 96720 + }, + { + "epoch": 28.143729997090485, + "grad_norm": 0.07440955936908722, + "learning_rate": 2.980203570987805e-06, + "loss": 0.0118, + "step": 96730 + }, + { + "epoch": 28.14663951120163, + "grad_norm": 0.09816870093345642, + "learning_rate": 2.977731113375329e-06, + "loss": 0.0106, + "step": 96740 + }, + { + "epoch": 28.149549025312773, + "grad_norm": 0.10693905502557755, + "learning_rate": 2.975259568774835e-06, + "loss": 0.0108, + "step": 96750 + }, + { + "epoch": 28.152458539423915, + "grad_norm": 0.20695947110652924, + "learning_rate": 2.972788937374024e-06, + "loss": 0.0122, + "step": 96760 + }, + { + "epoch": 28.15536805353506, + "grad_norm": 0.12052540481090546, + "learning_rate": 2.9703192193605194e-06, + "loss": 0.0126, + "step": 96770 + }, + { + "epoch": 28.158277567646202, + "grad_norm": 0.08253193646669388, + "learning_rate": 2.967850414921879e-06, + "loss": 0.0121, + "step": 96780 + }, + { + "epoch": 28.161187081757348, + "grad_norm": 0.08992686867713928, + "learning_rate": 2.9653825242455923e-06, + "loss": 0.0104, + "step": 96790 + }, + { + "epoch": 28.16409659586849, + "grad_norm": 0.09221556037664413, + "learning_rate": 2.9629155475190866e-06, + "loss": 0.0108, + "step": 96800 + }, + { + "epoch": 28.167006109979635, + "grad_norm": 0.08842688053846359, + "learning_rate": 2.9604494849297038e-06, + "loss": 0.0136, + "step": 96810 + }, + { + "epoch": 28.169915624090777, + "grad_norm": 0.09163209795951843, + "learning_rate": 2.9579843366647267e-06, + "loss": 0.0102, + "step": 96820 + }, + { + "epoch": 28.17282513820192, + "grad_norm": 0.12358071655035019, + "learning_rate": 2.955520102911375e-06, + "loss": 0.0121, + "step": 96830 + }, + { + "epoch": 28.175734652313064, + "grad_norm": 0.14167384803295135, + "learning_rate": 2.953056783856783e-06, + "loss": 0.0135, + "step": 96840 + }, + { + "epoch": 28.178644166424206, + "grad_norm": 0.11418069154024124, + "learning_rate": 2.950594379688023e-06, + "loss": 0.013, + "step": 96850 + }, + { + "epoch": 28.18155368053535, + "grad_norm": 0.06615874916315079, + "learning_rate": 2.948132890592099e-06, + "loss": 0.0092, + "step": 96860 + }, + { + "epoch": 28.184463194646494, + "grad_norm": 0.157173290848732, + "learning_rate": 2.9456723167559515e-06, + "loss": 0.0119, + "step": 96870 + }, + { + "epoch": 28.18737270875764, + "grad_norm": 0.1299230009317398, + "learning_rate": 2.9432126583664354e-06, + "loss": 0.0105, + "step": 96880 + }, + { + "epoch": 28.19028222286878, + "grad_norm": 0.08996094018220901, + "learning_rate": 2.940753915610355e-06, + "loss": 0.0127, + "step": 96890 + }, + { + "epoch": 28.193191736979923, + "grad_norm": 0.22014160454273224, + "learning_rate": 2.9382960886744264e-06, + "loss": 0.0127, + "step": 96900 + }, + { + "epoch": 28.19610125109107, + "grad_norm": 0.0880308449268341, + "learning_rate": 2.935839177745308e-06, + "loss": 0.0101, + "step": 96910 + }, + { + "epoch": 28.19901076520221, + "grad_norm": 0.12600280344486237, + "learning_rate": 2.933383183009594e-06, + "loss": 0.0117, + "step": 96920 + }, + { + "epoch": 28.201920279313356, + "grad_norm": 0.08913987129926682, + "learning_rate": 2.930928104653789e-06, + "loss": 0.0106, + "step": 96930 + }, + { + "epoch": 28.204829793424498, + "grad_norm": 0.08044512569904327, + "learning_rate": 2.92847394286435e-06, + "loss": 0.0102, + "step": 96940 + }, + { + "epoch": 28.207739307535643, + "grad_norm": 0.3798866271972656, + "learning_rate": 2.9260206978276444e-06, + "loss": 0.0141, + "step": 96950 + }, + { + "epoch": 28.210648821646785, + "grad_norm": 0.10851485282182693, + "learning_rate": 2.923568369729992e-06, + "loss": 0.0114, + "step": 96960 + }, + { + "epoch": 28.213558335757927, + "grad_norm": 0.1453690528869629, + "learning_rate": 2.9211169587576186e-06, + "loss": 0.0105, + "step": 96970 + }, + { + "epoch": 28.216467849869073, + "grad_norm": 0.08750668168067932, + "learning_rate": 2.918666465096699e-06, + "loss": 0.0115, + "step": 96980 + }, + { + "epoch": 28.219377363980215, + "grad_norm": 0.10711812227964401, + "learning_rate": 2.916216888933335e-06, + "loss": 0.0127, + "step": 96990 + }, + { + "epoch": 28.22228687809136, + "grad_norm": 0.1936221718788147, + "learning_rate": 2.9137682304535524e-06, + "loss": 0.012, + "step": 97000 + }, + { + "epoch": 28.225196392202502, + "grad_norm": 0.2614877223968506, + "learning_rate": 2.911320489843303e-06, + "loss": 0.0118, + "step": 97010 + }, + { + "epoch": 28.228105906313644, + "grad_norm": 0.13195118308067322, + "learning_rate": 2.9088736672884923e-06, + "loss": 0.0112, + "step": 97020 + }, + { + "epoch": 28.23101542042479, + "grad_norm": 0.11671222746372223, + "learning_rate": 2.9064277629749315e-06, + "loss": 0.0143, + "step": 97030 + }, + { + "epoch": 28.23392493453593, + "grad_norm": 0.1294061243534088, + "learning_rate": 2.903982777088368e-06, + "loss": 0.0102, + "step": 97040 + }, + { + "epoch": 28.236834448647077, + "grad_norm": 0.1323583424091339, + "learning_rate": 2.901538709814486e-06, + "loss": 0.0098, + "step": 97050 + }, + { + "epoch": 28.23974396275822, + "grad_norm": 0.09344936907291412, + "learning_rate": 2.8990955613388994e-06, + "loss": 0.0121, + "step": 97060 + }, + { + "epoch": 28.242653476869364, + "grad_norm": 0.11015602201223373, + "learning_rate": 2.896653331847145e-06, + "loss": 0.0114, + "step": 97070 + }, + { + "epoch": 28.245562990980506, + "grad_norm": 0.12129401415586472, + "learning_rate": 2.8942120215246916e-06, + "loss": 0.0101, + "step": 97080 + }, + { + "epoch": 28.248472505091648, + "grad_norm": 0.12382204085588455, + "learning_rate": 2.891771630556945e-06, + "loss": 0.0124, + "step": 97090 + }, + { + "epoch": 28.251382019202794, + "grad_norm": 0.09607204794883728, + "learning_rate": 2.889332159129238e-06, + "loss": 0.0089, + "step": 97100 + }, + { + "epoch": 28.254291533313936, + "grad_norm": 0.11401520669460297, + "learning_rate": 2.8868936074268322e-06, + "loss": 0.0097, + "step": 97110 + }, + { + "epoch": 28.25720104742508, + "grad_norm": 0.06479570269584656, + "learning_rate": 2.8844559756349073e-06, + "loss": 0.0105, + "step": 97120 + }, + { + "epoch": 28.260110561536223, + "grad_norm": 0.09403529763221741, + "learning_rate": 2.882019263938605e-06, + "loss": 0.0103, + "step": 97130 + }, + { + "epoch": 28.26302007564737, + "grad_norm": 0.10944997519254684, + "learning_rate": 2.8795834725229663e-06, + "loss": 0.0101, + "step": 97140 + }, + { + "epoch": 28.26592958975851, + "grad_norm": 0.11742119491100311, + "learning_rate": 2.8771486015729733e-06, + "loss": 0.0117, + "step": 97150 + }, + { + "epoch": 28.268839103869652, + "grad_norm": 0.23130394518375397, + "learning_rate": 2.874714651273537e-06, + "loss": 0.0115, + "step": 97160 + }, + { + "epoch": 28.271748617980798, + "grad_norm": 0.17706042528152466, + "learning_rate": 2.8722816218095095e-06, + "loss": 0.0127, + "step": 97170 + }, + { + "epoch": 28.27465813209194, + "grad_norm": 0.08700937777757645, + "learning_rate": 2.869849513365653e-06, + "loss": 0.0115, + "step": 97180 + }, + { + "epoch": 28.277567646203085, + "grad_norm": 0.1070399284362793, + "learning_rate": 2.8674183261266764e-06, + "loss": 0.0103, + "step": 97190 + }, + { + "epoch": 28.280477160314227, + "grad_norm": 0.10712640732526779, + "learning_rate": 2.864988060277207e-06, + "loss": 0.0105, + "step": 97200 + }, + { + "epoch": 28.283386674425373, + "grad_norm": 0.08952908217906952, + "learning_rate": 2.8625587160018136e-06, + "loss": 0.011, + "step": 97210 + }, + { + "epoch": 28.286296188536515, + "grad_norm": 0.22029472887516022, + "learning_rate": 2.8601302934849827e-06, + "loss": 0.0106, + "step": 97220 + }, + { + "epoch": 28.289205702647656, + "grad_norm": 0.18591037392616272, + "learning_rate": 2.8577027929111387e-06, + "loss": 0.0127, + "step": 97230 + }, + { + "epoch": 28.292115216758802, + "grad_norm": 0.1436394900083542, + "learning_rate": 2.8552762144646394e-06, + "loss": 0.0108, + "step": 97240 + }, + { + "epoch": 28.295024730869944, + "grad_norm": 0.09955525398254395, + "learning_rate": 2.8528505583297603e-06, + "loss": 0.0111, + "step": 97250 + }, + { + "epoch": 28.29793424498109, + "grad_norm": 0.10290397703647614, + "learning_rate": 2.8504258246907193e-06, + "loss": 0.0133, + "step": 97260 + }, + { + "epoch": 28.30084375909223, + "grad_norm": 0.12144531309604645, + "learning_rate": 2.8480020137316544e-06, + "loss": 0.011, + "step": 97270 + }, + { + "epoch": 28.303753273203377, + "grad_norm": 0.16139477491378784, + "learning_rate": 2.845579125636639e-06, + "loss": 0.0131, + "step": 97280 + }, + { + "epoch": 28.30666278731452, + "grad_norm": 0.11221969872713089, + "learning_rate": 2.843157160589682e-06, + "loss": 0.0101, + "step": 97290 + }, + { + "epoch": 28.30957230142566, + "grad_norm": 0.12737680971622467, + "learning_rate": 2.8407361187747113e-06, + "loss": 0.0121, + "step": 97300 + }, + { + "epoch": 28.312481815536806, + "grad_norm": 0.084770567715168, + "learning_rate": 2.838316000375582e-06, + "loss": 0.0101, + "step": 97310 + }, + { + "epoch": 28.315391329647948, + "grad_norm": 0.12830707430839539, + "learning_rate": 2.835896805576093e-06, + "loss": 0.0132, + "step": 97320 + }, + { + "epoch": 28.318300843759094, + "grad_norm": 0.09640057384967804, + "learning_rate": 2.8334785345599704e-06, + "loss": 0.0129, + "step": 97330 + }, + { + "epoch": 28.321210357870235, + "grad_norm": 0.07978063821792603, + "learning_rate": 2.831061187510856e-06, + "loss": 0.0098, + "step": 97340 + }, + { + "epoch": 28.324119871981377, + "grad_norm": 0.14277233183383942, + "learning_rate": 2.8286447646123376e-06, + "loss": 0.0126, + "step": 97350 + }, + { + "epoch": 28.327029386092523, + "grad_norm": 0.13635756075382233, + "learning_rate": 2.8262292660479283e-06, + "loss": 0.0134, + "step": 97360 + }, + { + "epoch": 28.329938900203665, + "grad_norm": 0.1713048815727234, + "learning_rate": 2.8238146920010667e-06, + "loss": 0.0105, + "step": 97370 + }, + { + "epoch": 28.33284841431481, + "grad_norm": 0.1520623415708542, + "learning_rate": 2.8214010426551207e-06, + "loss": 0.0125, + "step": 97380 + }, + { + "epoch": 28.335757928425952, + "grad_norm": 0.13453158736228943, + "learning_rate": 2.8189883181933922e-06, + "loss": 0.0117, + "step": 97390 + }, + { + "epoch": 28.338667442537098, + "grad_norm": 0.08837773650884628, + "learning_rate": 2.8165765187991167e-06, + "loss": 0.0102, + "step": 97400 + }, + { + "epoch": 28.34157695664824, + "grad_norm": 0.06950005143880844, + "learning_rate": 2.814165644655452e-06, + "loss": 0.0102, + "step": 97410 + }, + { + "epoch": 28.34448647075938, + "grad_norm": 0.15676435828208923, + "learning_rate": 2.8117556959454787e-06, + "loss": 0.0102, + "step": 97420 + }, + { + "epoch": 28.347395984870527, + "grad_norm": 0.11227737367153168, + "learning_rate": 2.809346672852234e-06, + "loss": 0.0103, + "step": 97430 + }, + { + "epoch": 28.35030549898167, + "grad_norm": 0.10188087075948715, + "learning_rate": 2.8069385755586572e-06, + "loss": 0.0118, + "step": 97440 + }, + { + "epoch": 28.353215013092814, + "grad_norm": 0.11687441915273666, + "learning_rate": 2.804531404247625e-06, + "loss": 0.0116, + "step": 97450 + }, + { + "epoch": 28.356124527203956, + "grad_norm": 0.08592475950717926, + "learning_rate": 2.8021251591019486e-06, + "loss": 0.0097, + "step": 97460 + }, + { + "epoch": 28.359034041315102, + "grad_norm": 0.13374298810958862, + "learning_rate": 2.7997198403043724e-06, + "loss": 0.0103, + "step": 97470 + }, + { + "epoch": 28.361943555426244, + "grad_norm": 0.09212581813335419, + "learning_rate": 2.7973154480375545e-06, + "loss": 0.0123, + "step": 97480 + }, + { + "epoch": 28.364853069537386, + "grad_norm": 0.27165988087654114, + "learning_rate": 2.794911982484101e-06, + "loss": 0.0103, + "step": 97490 + }, + { + "epoch": 28.36776258364853, + "grad_norm": 0.178179532289505, + "learning_rate": 2.792509443826533e-06, + "loss": 0.0115, + "step": 97500 + }, + { + "epoch": 28.370672097759673, + "grad_norm": 0.5642987489700317, + "learning_rate": 2.790107832247314e-06, + "loss": 0.0099, + "step": 97510 + }, + { + "epoch": 28.37358161187082, + "grad_norm": 0.1536235511302948, + "learning_rate": 2.7877071479288248e-06, + "loss": 0.0109, + "step": 97520 + }, + { + "epoch": 28.37649112598196, + "grad_norm": 0.14793570339679718, + "learning_rate": 2.785307391053381e-06, + "loss": 0.0111, + "step": 97530 + }, + { + "epoch": 28.379400640093106, + "grad_norm": 0.09213881939649582, + "learning_rate": 2.7829085618032355e-06, + "loss": 0.0115, + "step": 97540 + }, + { + "epoch": 28.382310154204248, + "grad_norm": 0.08098451048135757, + "learning_rate": 2.780510660360556e-06, + "loss": 0.0109, + "step": 97550 + }, + { + "epoch": 28.38521966831539, + "grad_norm": 0.14281107485294342, + "learning_rate": 2.7781136869074535e-06, + "loss": 0.0113, + "step": 97560 + }, + { + "epoch": 28.388129182426535, + "grad_norm": 0.11939889937639236, + "learning_rate": 2.775717641625954e-06, + "loss": 0.0116, + "step": 97570 + }, + { + "epoch": 28.391038696537677, + "grad_norm": 0.08694462478160858, + "learning_rate": 2.7733225246980315e-06, + "loss": 0.0101, + "step": 97580 + }, + { + "epoch": 28.393948210648823, + "grad_norm": 0.08746728301048279, + "learning_rate": 2.770928336305571e-06, + "loss": 0.0119, + "step": 97590 + }, + { + "epoch": 28.396857724759965, + "grad_norm": 0.11682799458503723, + "learning_rate": 2.7685350766304e-06, + "loss": 0.0107, + "step": 97600 + }, + { + "epoch": 28.39976723887111, + "grad_norm": 0.13480263948440552, + "learning_rate": 2.766142745854267e-06, + "loss": 0.0109, + "step": 97610 + }, + { + "epoch": 28.402676752982252, + "grad_norm": 0.09539345651865005, + "learning_rate": 2.763751344158856e-06, + "loss": 0.0093, + "step": 97620 + }, + { + "epoch": 28.405586267093394, + "grad_norm": 0.09372766315937042, + "learning_rate": 2.7613608717257827e-06, + "loss": 0.0095, + "step": 97630 + }, + { + "epoch": 28.40849578120454, + "grad_norm": 0.06015473231673241, + "learning_rate": 2.7589713287365794e-06, + "loss": 0.0095, + "step": 97640 + }, + { + "epoch": 28.41140529531568, + "grad_norm": 0.07191997766494751, + "learning_rate": 2.756582715372719e-06, + "loss": 0.0129, + "step": 97650 + }, + { + "epoch": 28.414314809426827, + "grad_norm": 0.1006719097495079, + "learning_rate": 2.7541950318156056e-06, + "loss": 0.0111, + "step": 97660 + }, + { + "epoch": 28.41722432353797, + "grad_norm": 0.09715109318494797, + "learning_rate": 2.7518082782465635e-06, + "loss": 0.012, + "step": 97670 + }, + { + "epoch": 28.42013383764911, + "grad_norm": 0.12458264827728271, + "learning_rate": 2.74942245484685e-06, + "loss": 0.0096, + "step": 97680 + }, + { + "epoch": 28.423043351760256, + "grad_norm": 0.11561909317970276, + "learning_rate": 2.747037561797653e-06, + "loss": 0.0099, + "step": 97690 + }, + { + "epoch": 28.4259528658714, + "grad_norm": 0.095987968146801, + "learning_rate": 2.744653599280094e-06, + "loss": 0.0098, + "step": 97700 + }, + { + "epoch": 28.428862379982544, + "grad_norm": 0.1920228898525238, + "learning_rate": 2.7422705674752156e-06, + "loss": 0.0126, + "step": 97710 + }, + { + "epoch": 28.431771894093686, + "grad_norm": 0.09745332598686218, + "learning_rate": 2.7398884665639877e-06, + "loss": 0.0106, + "step": 97720 + }, + { + "epoch": 28.43468140820483, + "grad_norm": 0.10342087596654892, + "learning_rate": 2.7375072967273275e-06, + "loss": 0.0117, + "step": 97730 + }, + { + "epoch": 28.437590922315973, + "grad_norm": 0.10566645860671997, + "learning_rate": 2.735127058146063e-06, + "loss": 0.0097, + "step": 97740 + }, + { + "epoch": 28.440500436427115, + "grad_norm": 0.09446866810321808, + "learning_rate": 2.7327477510009536e-06, + "loss": 0.0111, + "step": 97750 + }, + { + "epoch": 28.44340995053826, + "grad_norm": 0.10791753232479095, + "learning_rate": 2.730369375472696e-06, + "loss": 0.0105, + "step": 97760 + }, + { + "epoch": 28.446319464649402, + "grad_norm": 0.08652960509061813, + "learning_rate": 2.727991931741916e-06, + "loss": 0.009, + "step": 97770 + }, + { + "epoch": 28.449228978760548, + "grad_norm": 0.12296925485134125, + "learning_rate": 2.725615419989159e-06, + "loss": 0.0126, + "step": 97780 + }, + { + "epoch": 28.45213849287169, + "grad_norm": 0.17461329698562622, + "learning_rate": 2.7232398403949043e-06, + "loss": 0.014, + "step": 97790 + }, + { + "epoch": 28.455048006982835, + "grad_norm": 0.10400432348251343, + "learning_rate": 2.720865193139564e-06, + "loss": 0.0091, + "step": 97800 + }, + { + "epoch": 28.457957521093977, + "grad_norm": 0.10892391949892044, + "learning_rate": 2.7184914784034804e-06, + "loss": 0.0094, + "step": 97810 + }, + { + "epoch": 28.46086703520512, + "grad_norm": 0.08354543894529343, + "learning_rate": 2.7161186963669145e-06, + "loss": 0.0121, + "step": 97820 + }, + { + "epoch": 28.463776549316265, + "grad_norm": 0.10773086547851562, + "learning_rate": 2.713746847210067e-06, + "loss": 0.0128, + "step": 97830 + }, + { + "epoch": 28.466686063427407, + "grad_norm": 0.14034081995487213, + "learning_rate": 2.7113759311130688e-06, + "loss": 0.0112, + "step": 97840 + }, + { + "epoch": 28.469595577538552, + "grad_norm": 0.07994251698255539, + "learning_rate": 2.709005948255966e-06, + "loss": 0.0085, + "step": 97850 + }, + { + "epoch": 28.472505091649694, + "grad_norm": 0.15023356676101685, + "learning_rate": 2.706636898818753e-06, + "loss": 0.0089, + "step": 97860 + }, + { + "epoch": 28.47541460576084, + "grad_norm": 0.10449380427598953, + "learning_rate": 2.7042687829813323e-06, + "loss": 0.0119, + "step": 97870 + }, + { + "epoch": 28.47832411987198, + "grad_norm": 0.130879208445549, + "learning_rate": 2.7019016009235588e-06, + "loss": 0.0106, + "step": 97880 + }, + { + "epoch": 28.481233633983123, + "grad_norm": 0.12427414953708649, + "learning_rate": 2.699535352825193e-06, + "loss": 0.0104, + "step": 97890 + }, + { + "epoch": 28.48414314809427, + "grad_norm": 0.11723575741052628, + "learning_rate": 2.6971700388659452e-06, + "loss": 0.0122, + "step": 97900 + }, + { + "epoch": 28.48705266220541, + "grad_norm": 0.10459046065807343, + "learning_rate": 2.694805659225438e-06, + "loss": 0.0112, + "step": 97910 + }, + { + "epoch": 28.489962176316556, + "grad_norm": 0.08155040442943573, + "learning_rate": 2.692442214083235e-06, + "loss": 0.0095, + "step": 97920 + }, + { + "epoch": 28.492871690427698, + "grad_norm": 0.08515691012144089, + "learning_rate": 2.690079703618825e-06, + "loss": 0.0113, + "step": 97930 + }, + { + "epoch": 28.495781204538844, + "grad_norm": 0.11430877447128296, + "learning_rate": 2.6877181280116203e-06, + "loss": 0.0092, + "step": 97940 + }, + { + "epoch": 28.498690718649986, + "grad_norm": 0.07788780331611633, + "learning_rate": 2.6853574874409736e-06, + "loss": 0.0087, + "step": 97950 + }, + { + "epoch": 28.501600232761128, + "grad_norm": 0.14842405915260315, + "learning_rate": 2.682997782086152e-06, + "loss": 0.012, + "step": 97960 + }, + { + "epoch": 28.504509746872273, + "grad_norm": 0.09185489267110825, + "learning_rate": 2.680639012126366e-06, + "loss": 0.0118, + "step": 97970 + }, + { + "epoch": 28.507419260983415, + "grad_norm": 0.22887377440929413, + "learning_rate": 2.678281177740744e-06, + "loss": 0.0121, + "step": 97980 + }, + { + "epoch": 28.51032877509456, + "grad_norm": 0.10567469894886017, + "learning_rate": 2.6759242791083518e-06, + "loss": 0.0103, + "step": 97990 + }, + { + "epoch": 28.513238289205702, + "grad_norm": 0.11717712134122849, + "learning_rate": 2.673568316408181e-06, + "loss": 0.0105, + "step": 98000 + }, + { + "epoch": 28.516147803316848, + "grad_norm": 0.10904695838689804, + "learning_rate": 2.671213289819149e-06, + "loss": 0.0119, + "step": 98010 + }, + { + "epoch": 28.51905731742799, + "grad_norm": 0.29598185420036316, + "learning_rate": 2.668859199520097e-06, + "loss": 0.0124, + "step": 98020 + }, + { + "epoch": 28.52196683153913, + "grad_norm": 0.09971896559000015, + "learning_rate": 2.666506045689818e-06, + "loss": 0.0128, + "step": 98030 + }, + { + "epoch": 28.524876345650277, + "grad_norm": 0.09877574443817139, + "learning_rate": 2.6641538285070107e-06, + "loss": 0.0114, + "step": 98040 + }, + { + "epoch": 28.52778585976142, + "grad_norm": 0.1170063465833664, + "learning_rate": 2.6618025481503077e-06, + "loss": 0.0109, + "step": 98050 + }, + { + "epoch": 28.530695373872565, + "grad_norm": 0.09255421161651611, + "learning_rate": 2.6594522047982756e-06, + "loss": 0.0103, + "step": 98060 + }, + { + "epoch": 28.533604887983707, + "grad_norm": 0.08552636951208115, + "learning_rate": 2.657102798629413e-06, + "loss": 0.0107, + "step": 98070 + }, + { + "epoch": 28.53651440209485, + "grad_norm": 0.12302565574645996, + "learning_rate": 2.654754329822135e-06, + "loss": 0.0092, + "step": 98080 + }, + { + "epoch": 28.539423916205994, + "grad_norm": 0.09225117415189743, + "learning_rate": 2.652406798554791e-06, + "loss": 0.0104, + "step": 98090 + }, + { + "epoch": 28.542333430317136, + "grad_norm": 0.10911740362644196, + "learning_rate": 2.650060205005664e-06, + "loss": 0.0096, + "step": 98100 + }, + { + "epoch": 28.54524294442828, + "grad_norm": 0.0957140177488327, + "learning_rate": 2.6477145493529652e-06, + "loss": 0.012, + "step": 98110 + }, + { + "epoch": 28.548152458539423, + "grad_norm": 0.09034126996994019, + "learning_rate": 2.6453698317748247e-06, + "loss": 0.0097, + "step": 98120 + }, + { + "epoch": 28.55106197265057, + "grad_norm": 0.10008101910352707, + "learning_rate": 2.6430260524493116e-06, + "loss": 0.0107, + "step": 98130 + }, + { + "epoch": 28.55397148676171, + "grad_norm": 0.07760139554738998, + "learning_rate": 2.6406832115544247e-06, + "loss": 0.0101, + "step": 98140 + }, + { + "epoch": 28.556881000872853, + "grad_norm": 0.08735030144453049, + "learning_rate": 2.6383413092680843e-06, + "loss": 0.0084, + "step": 98150 + }, + { + "epoch": 28.559790514983998, + "grad_norm": 0.1555217206478119, + "learning_rate": 2.6360003457681387e-06, + "loss": 0.0103, + "step": 98160 + }, + { + "epoch": 28.56270002909514, + "grad_norm": 0.083661749958992, + "learning_rate": 2.6336603212323693e-06, + "loss": 0.0104, + "step": 98170 + }, + { + "epoch": 28.565609543206286, + "grad_norm": 0.12828482687473297, + "learning_rate": 2.631321235838493e-06, + "loss": 0.0127, + "step": 98180 + }, + { + "epoch": 28.568519057317427, + "grad_norm": 0.09268027544021606, + "learning_rate": 2.6289830897641386e-06, + "loss": 0.0122, + "step": 98190 + }, + { + "epoch": 28.571428571428573, + "grad_norm": 0.15208229422569275, + "learning_rate": 2.6266458831868797e-06, + "loss": 0.0099, + "step": 98200 + }, + { + "epoch": 28.574338085539715, + "grad_norm": 0.11315120756626129, + "learning_rate": 2.624309616284207e-06, + "loss": 0.0111, + "step": 98210 + }, + { + "epoch": 28.577247599650857, + "grad_norm": 0.09908509254455566, + "learning_rate": 2.621974289233545e-06, + "loss": 0.0097, + "step": 98220 + }, + { + "epoch": 28.580157113762002, + "grad_norm": 0.1012888178229332, + "learning_rate": 2.619639902212251e-06, + "loss": 0.009, + "step": 98230 + }, + { + "epoch": 28.583066627873144, + "grad_norm": 0.09000766277313232, + "learning_rate": 2.6173064553976016e-06, + "loss": 0.0116, + "step": 98240 + }, + { + "epoch": 28.58597614198429, + "grad_norm": 0.10230209678411484, + "learning_rate": 2.6149739489668113e-06, + "loss": 0.0106, + "step": 98250 + }, + { + "epoch": 28.58888565609543, + "grad_norm": 0.0878186821937561, + "learning_rate": 2.6126423830970114e-06, + "loss": 0.0115, + "step": 98260 + }, + { + "epoch": 28.591795170206577, + "grad_norm": 0.08747760206460953, + "learning_rate": 2.6103117579652758e-06, + "loss": 0.0103, + "step": 98270 + }, + { + "epoch": 28.59470468431772, + "grad_norm": 0.10793249309062958, + "learning_rate": 2.607982073748596e-06, + "loss": 0.0107, + "step": 98280 + }, + { + "epoch": 28.59761419842886, + "grad_norm": 0.12229033559560776, + "learning_rate": 2.6056533306238955e-06, + "loss": 0.0104, + "step": 98290 + }, + { + "epoch": 28.600523712540006, + "grad_norm": 0.1567593216896057, + "learning_rate": 2.603325528768033e-06, + "loss": 0.0135, + "step": 98300 + }, + { + "epoch": 28.60343322665115, + "grad_norm": 0.12385322898626328, + "learning_rate": 2.6009986683577867e-06, + "loss": 0.0102, + "step": 98310 + }, + { + "epoch": 28.606342740762294, + "grad_norm": 0.10537239909172058, + "learning_rate": 2.598672749569861e-06, + "loss": 0.0093, + "step": 98320 + }, + { + "epoch": 28.609252254873436, + "grad_norm": 0.09883871674537659, + "learning_rate": 2.596347772580897e-06, + "loss": 0.0125, + "step": 98330 + }, + { + "epoch": 28.612161768984578, + "grad_norm": 0.11696377396583557, + "learning_rate": 2.5940237375674673e-06, + "loss": 0.0105, + "step": 98340 + }, + { + "epoch": 28.615071283095723, + "grad_norm": 0.11016026884317398, + "learning_rate": 2.5917006447060595e-06, + "loss": 0.013, + "step": 98350 + }, + { + "epoch": 28.617980797206865, + "grad_norm": 0.12948860228061676, + "learning_rate": 2.5893784941730995e-06, + "loss": 0.012, + "step": 98360 + }, + { + "epoch": 28.62089031131801, + "grad_norm": 0.05309295281767845, + "learning_rate": 2.5870572861449443e-06, + "loss": 0.0101, + "step": 98370 + }, + { + "epoch": 28.623799825429153, + "grad_norm": 0.11482275277376175, + "learning_rate": 2.5847370207978687e-06, + "loss": 0.009, + "step": 98380 + }, + { + "epoch": 28.626709339540298, + "grad_norm": 0.09720452129840851, + "learning_rate": 2.5824176983080804e-06, + "loss": 0.0105, + "step": 98390 + }, + { + "epoch": 28.62961885365144, + "grad_norm": 0.08250757306814194, + "learning_rate": 2.580099318851717e-06, + "loss": 0.0109, + "step": 98400 + }, + { + "epoch": 28.632528367762582, + "grad_norm": 0.113948754966259, + "learning_rate": 2.577781882604852e-06, + "loss": 0.0128, + "step": 98410 + }, + { + "epoch": 28.635437881873727, + "grad_norm": 0.1073017418384552, + "learning_rate": 2.575465389743472e-06, + "loss": 0.0119, + "step": 98420 + }, + { + "epoch": 28.63834739598487, + "grad_norm": 0.09532332420349121, + "learning_rate": 2.5731498404434945e-06, + "loss": 0.0106, + "step": 98430 + }, + { + "epoch": 28.641256910096015, + "grad_norm": 0.23106525838375092, + "learning_rate": 2.5708352348807835e-06, + "loss": 0.0106, + "step": 98440 + }, + { + "epoch": 28.644166424207157, + "grad_norm": 0.1134149432182312, + "learning_rate": 2.5685215732311103e-06, + "loss": 0.0095, + "step": 98450 + }, + { + "epoch": 28.647075938318302, + "grad_norm": 0.08352121710777283, + "learning_rate": 2.5662088556701807e-06, + "loss": 0.0091, + "step": 98460 + }, + { + "epoch": 28.649985452429444, + "grad_norm": 0.10607556998729706, + "learning_rate": 2.5638970823736312e-06, + "loss": 0.011, + "step": 98470 + }, + { + "epoch": 28.652894966540586, + "grad_norm": 0.08597410470247269, + "learning_rate": 2.5615862535170316e-06, + "loss": 0.0108, + "step": 98480 + }, + { + "epoch": 28.65580448065173, + "grad_norm": 0.10640964657068253, + "learning_rate": 2.559276369275866e-06, + "loss": 0.0117, + "step": 98490 + }, + { + "epoch": 28.658713994762874, + "grad_norm": 0.06963127851486206, + "learning_rate": 2.5569674298255625e-06, + "loss": 0.0112, + "step": 98500 + }, + { + "epoch": 28.66162350887402, + "grad_norm": 0.09400637447834015, + "learning_rate": 2.5546594353414625e-06, + "loss": 0.0108, + "step": 98510 + }, + { + "epoch": 28.66453302298516, + "grad_norm": 0.10921163856983185, + "learning_rate": 2.552352385998848e-06, + "loss": 0.0112, + "step": 98520 + }, + { + "epoch": 28.667442537096306, + "grad_norm": 0.12310309708118439, + "learning_rate": 2.5500462819729213e-06, + "loss": 0.0112, + "step": 98530 + }, + { + "epoch": 28.67035205120745, + "grad_norm": 0.1397581696510315, + "learning_rate": 2.5477411234388147e-06, + "loss": 0.012, + "step": 98540 + }, + { + "epoch": 28.67326156531859, + "grad_norm": 0.09340006858110428, + "learning_rate": 2.5454369105715965e-06, + "loss": 0.0112, + "step": 98550 + }, + { + "epoch": 28.676171079429736, + "grad_norm": 0.08621671050786972, + "learning_rate": 2.5431336435462494e-06, + "loss": 0.0107, + "step": 98560 + }, + { + "epoch": 28.679080593540878, + "grad_norm": 0.12417156994342804, + "learning_rate": 2.5408313225376955e-06, + "loss": 0.0111, + "step": 98570 + }, + { + "epoch": 28.681990107652023, + "grad_norm": 0.0848313719034195, + "learning_rate": 2.538529947720777e-06, + "loss": 0.0115, + "step": 98580 + }, + { + "epoch": 28.684899621763165, + "grad_norm": 0.1921319216489792, + "learning_rate": 2.53622951927027e-06, + "loss": 0.0118, + "step": 98590 + }, + { + "epoch": 28.68780913587431, + "grad_norm": 0.07869099825620651, + "learning_rate": 2.533930037360882e-06, + "loss": 0.0105, + "step": 98600 + }, + { + "epoch": 28.690718649985453, + "grad_norm": 0.09091523289680481, + "learning_rate": 2.5316315021672384e-06, + "loss": 0.0108, + "step": 98610 + }, + { + "epoch": 28.693628164096594, + "grad_norm": 0.09435641765594482, + "learning_rate": 2.5293339138638937e-06, + "loss": 0.0103, + "step": 98620 + }, + { + "epoch": 28.69653767820774, + "grad_norm": 0.10650788247585297, + "learning_rate": 2.5270372726253392e-06, + "loss": 0.0106, + "step": 98630 + }, + { + "epoch": 28.699447192318882, + "grad_norm": 0.08433486521244049, + "learning_rate": 2.5247415786259938e-06, + "loss": 0.0108, + "step": 98640 + }, + { + "epoch": 28.702356706430027, + "grad_norm": 0.11160383373498917, + "learning_rate": 2.5224468320401907e-06, + "loss": 0.0104, + "step": 98650 + }, + { + "epoch": 28.70526622054117, + "grad_norm": 0.07961607724428177, + "learning_rate": 2.5201530330422075e-06, + "loss": 0.0124, + "step": 98660 + }, + { + "epoch": 28.708175734652315, + "grad_norm": 0.12794265151023865, + "learning_rate": 2.517860181806245e-06, + "loss": 0.0112, + "step": 98670 + }, + { + "epoch": 28.711085248763457, + "grad_norm": 0.13509564101696014, + "learning_rate": 2.5155682785064266e-06, + "loss": 0.0118, + "step": 98680 + }, + { + "epoch": 28.7139947628746, + "grad_norm": 0.0991855338215828, + "learning_rate": 2.5132773233168006e-06, + "loss": 0.0109, + "step": 98690 + }, + { + "epoch": 28.716904276985744, + "grad_norm": 0.10036944597959518, + "learning_rate": 2.5109873164113623e-06, + "loss": 0.0126, + "step": 98700 + }, + { + "epoch": 28.719813791096886, + "grad_norm": 0.14113402366638184, + "learning_rate": 2.5086982579640177e-06, + "loss": 0.0109, + "step": 98710 + }, + { + "epoch": 28.72272330520803, + "grad_norm": 0.09047998487949371, + "learning_rate": 2.5064101481486062e-06, + "loss": 0.0096, + "step": 98720 + }, + { + "epoch": 28.725632819319173, + "grad_norm": 0.12691077589988708, + "learning_rate": 2.5041229871388854e-06, + "loss": 0.0135, + "step": 98730 + }, + { + "epoch": 28.72854233343032, + "grad_norm": 0.09234873205423355, + "learning_rate": 2.5018367751085647e-06, + "loss": 0.0098, + "step": 98740 + }, + { + "epoch": 28.73145184754146, + "grad_norm": 0.11564215272665024, + "learning_rate": 2.4995515122312616e-06, + "loss": 0.0103, + "step": 98750 + }, + { + "epoch": 28.734361361652603, + "grad_norm": 0.09374123066663742, + "learning_rate": 2.497267198680521e-06, + "loss": 0.0113, + "step": 98760 + }, + { + "epoch": 28.73727087576375, + "grad_norm": 0.1096271201968193, + "learning_rate": 2.494983834629826e-06, + "loss": 0.0097, + "step": 98770 + }, + { + "epoch": 28.74018038987489, + "grad_norm": 0.13074736297130585, + "learning_rate": 2.4927014202525856e-06, + "loss": 0.0098, + "step": 98780 + }, + { + "epoch": 28.743089903986036, + "grad_norm": 0.09975916147232056, + "learning_rate": 2.490419955722133e-06, + "loss": 0.0115, + "step": 98790 + }, + { + "epoch": 28.745999418097178, + "grad_norm": 0.16687904298305511, + "learning_rate": 2.48813944121172e-06, + "loss": 0.0108, + "step": 98800 + }, + { + "epoch": 28.74890893220832, + "grad_norm": 0.1188945472240448, + "learning_rate": 2.4858598768945523e-06, + "loss": 0.0101, + "step": 98810 + }, + { + "epoch": 28.751818446319465, + "grad_norm": 0.10018423944711685, + "learning_rate": 2.483581262943742e-06, + "loss": 0.011, + "step": 98820 + }, + { + "epoch": 28.754727960430607, + "grad_norm": 0.11326644569635391, + "learning_rate": 2.481303599532329e-06, + "loss": 0.0136, + "step": 98830 + }, + { + "epoch": 28.757637474541752, + "grad_norm": 0.1016308069229126, + "learning_rate": 2.4790268868332935e-06, + "loss": 0.0102, + "step": 98840 + }, + { + "epoch": 28.760546988652894, + "grad_norm": 0.09539709240198135, + "learning_rate": 2.476751125019536e-06, + "loss": 0.0102, + "step": 98850 + }, + { + "epoch": 28.76345650276404, + "grad_norm": 0.12865202128887177, + "learning_rate": 2.4744763142638817e-06, + "loss": 0.01, + "step": 98860 + }, + { + "epoch": 28.766366016875182, + "grad_norm": 0.11082407832145691, + "learning_rate": 2.4722024547390926e-06, + "loss": 0.0119, + "step": 98870 + }, + { + "epoch": 28.769275530986324, + "grad_norm": 0.10277251899242401, + "learning_rate": 2.4699295466178496e-06, + "loss": 0.0101, + "step": 98880 + }, + { + "epoch": 28.77218504509747, + "grad_norm": 0.11792582273483276, + "learning_rate": 2.4676575900727675e-06, + "loss": 0.0109, + "step": 98890 + }, + { + "epoch": 28.77509455920861, + "grad_norm": 0.08815564960241318, + "learning_rate": 2.4653865852763823e-06, + "loss": 0.0122, + "step": 98900 + }, + { + "epoch": 28.778004073319757, + "grad_norm": 0.09326225519180298, + "learning_rate": 2.463116532401166e-06, + "loss": 0.0116, + "step": 98910 + }, + { + "epoch": 28.7809135874309, + "grad_norm": 0.14395244419574738, + "learning_rate": 2.4608474316195144e-06, + "loss": 0.0106, + "step": 98920 + }, + { + "epoch": 28.783823101542044, + "grad_norm": 0.0867842584848404, + "learning_rate": 2.4585792831037463e-06, + "loss": 0.0108, + "step": 98930 + }, + { + "epoch": 28.786732615653186, + "grad_norm": 0.15091918408870697, + "learning_rate": 2.4563120870261195e-06, + "loss": 0.0127, + "step": 98940 + }, + { + "epoch": 28.789642129764328, + "grad_norm": 0.08363914489746094, + "learning_rate": 2.454045843558804e-06, + "loss": 0.0109, + "step": 98950 + }, + { + "epoch": 28.792551643875473, + "grad_norm": 0.11258859187364578, + "learning_rate": 2.4517805528739107e-06, + "loss": 0.0105, + "step": 98960 + }, + { + "epoch": 28.795461157986615, + "grad_norm": 0.14858020842075348, + "learning_rate": 2.449516215143478e-06, + "loss": 0.0102, + "step": 98970 + }, + { + "epoch": 28.79837067209776, + "grad_norm": 0.06824581325054169, + "learning_rate": 2.447252830539463e-06, + "loss": 0.0101, + "step": 98980 + }, + { + "epoch": 28.801280186208903, + "grad_norm": 0.10020966082811356, + "learning_rate": 2.44499039923375e-06, + "loss": 0.0088, + "step": 98990 + }, + { + "epoch": 28.804189700320048, + "grad_norm": 0.10217002034187317, + "learning_rate": 2.4427289213981602e-06, + "loss": 0.0117, + "step": 99000 + }, + { + "epoch": 28.80709921443119, + "grad_norm": 0.07805261760950089, + "learning_rate": 2.440468397204442e-06, + "loss": 0.012, + "step": 99010 + }, + { + "epoch": 28.810008728542332, + "grad_norm": 0.07987900823354721, + "learning_rate": 2.438208826824259e-06, + "loss": 0.0114, + "step": 99020 + }, + { + "epoch": 28.812918242653478, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.435950210429217e-06, + "loss": 0.0108, + "step": 99030 + }, + { + "epoch": 28.81582775676462, + "grad_norm": 0.11924725025892258, + "learning_rate": 2.4336925481908435e-06, + "loss": 0.0111, + "step": 99040 + }, + { + "epoch": 28.818737270875765, + "grad_norm": 0.09126931428909302, + "learning_rate": 2.43143584028059e-06, + "loss": 0.0093, + "step": 99050 + }, + { + "epoch": 28.821646784986907, + "grad_norm": 0.07232232391834259, + "learning_rate": 2.4291800868698354e-06, + "loss": 0.0115, + "step": 99060 + }, + { + "epoch": 28.82455629909805, + "grad_norm": 0.11351070553064346, + "learning_rate": 2.4269252881298936e-06, + "loss": 0.0106, + "step": 99070 + }, + { + "epoch": 28.827465813209194, + "grad_norm": 0.15333552658557892, + "learning_rate": 2.4246714442320057e-06, + "loss": 0.0107, + "step": 99080 + }, + { + "epoch": 28.830375327320336, + "grad_norm": 0.16162265837192535, + "learning_rate": 2.422418555347329e-06, + "loss": 0.0085, + "step": 99090 + }, + { + "epoch": 28.83328484143148, + "grad_norm": 0.10288070887327194, + "learning_rate": 2.4201666216469533e-06, + "loss": 0.011, + "step": 99100 + }, + { + "epoch": 28.836194355542624, + "grad_norm": 0.1820330023765564, + "learning_rate": 2.41791564330191e-06, + "loss": 0.0133, + "step": 99110 + }, + { + "epoch": 28.83910386965377, + "grad_norm": 0.09438538551330566, + "learning_rate": 2.4156656204831385e-06, + "loss": 0.0105, + "step": 99120 + }, + { + "epoch": 28.84201338376491, + "grad_norm": 0.0913604274392128, + "learning_rate": 2.4134165533615097e-06, + "loss": 0.0102, + "step": 99130 + }, + { + "epoch": 28.844922897876053, + "grad_norm": 0.1055363342165947, + "learning_rate": 2.4111684421078294e-06, + "loss": 0.0109, + "step": 99140 + }, + { + "epoch": 28.8478324119872, + "grad_norm": 0.11891163885593414, + "learning_rate": 2.4089212868928323e-06, + "loss": 0.0119, + "step": 99150 + }, + { + "epoch": 28.85074192609834, + "grad_norm": 0.0696154460310936, + "learning_rate": 2.4066750878871668e-06, + "loss": 0.0112, + "step": 99160 + }, + { + "epoch": 28.853651440209486, + "grad_norm": 0.0779939815402031, + "learning_rate": 2.4044298452614185e-06, + "loss": 0.0109, + "step": 99170 + }, + { + "epoch": 28.856560954320628, + "grad_norm": 0.08058398962020874, + "learning_rate": 2.4021855591860985e-06, + "loss": 0.0093, + "step": 99180 + }, + { + "epoch": 28.859470468431773, + "grad_norm": 0.1437763273715973, + "learning_rate": 2.3999422298316515e-06, + "loss": 0.0101, + "step": 99190 + }, + { + "epoch": 28.862379982542915, + "grad_norm": 0.1369076669216156, + "learning_rate": 2.3976998573684353e-06, + "loss": 0.0122, + "step": 99200 + }, + { + "epoch": 28.865289496654057, + "grad_norm": 0.12028560042381287, + "learning_rate": 2.3954584419667463e-06, + "loss": 0.0095, + "step": 99210 + }, + { + "epoch": 28.868199010765203, + "grad_norm": 0.10009239614009857, + "learning_rate": 2.3932179837968093e-06, + "loss": 0.0124, + "step": 99220 + }, + { + "epoch": 28.871108524876345, + "grad_norm": 0.16398055851459503, + "learning_rate": 2.3909784830287666e-06, + "loss": 0.013, + "step": 99230 + }, + { + "epoch": 28.87401803898749, + "grad_norm": 0.10879532247781754, + "learning_rate": 2.3887399398326976e-06, + "loss": 0.0126, + "step": 99240 + }, + { + "epoch": 28.876927553098632, + "grad_norm": 0.17434367537498474, + "learning_rate": 2.3865023543786007e-06, + "loss": 0.0119, + "step": 99250 + }, + { + "epoch": 28.879837067209778, + "grad_norm": 0.12708435952663422, + "learning_rate": 2.384265726836412e-06, + "loss": 0.0117, + "step": 99260 + }, + { + "epoch": 28.88274658132092, + "grad_norm": 0.09088437259197235, + "learning_rate": 2.382030057375981e-06, + "loss": 0.0096, + "step": 99270 + }, + { + "epoch": 28.88565609543206, + "grad_norm": 0.08715971559286118, + "learning_rate": 2.379795346167099e-06, + "loss": 0.0105, + "step": 99280 + }, + { + "epoch": 28.888565609543207, + "grad_norm": 0.08041825890541077, + "learning_rate": 2.3775615933794697e-06, + "loss": 0.01, + "step": 99290 + }, + { + "epoch": 28.89147512365435, + "grad_norm": 0.1375250518321991, + "learning_rate": 2.375328799182739e-06, + "loss": 0.0137, + "step": 99300 + }, + { + "epoch": 28.894384637765494, + "grad_norm": 0.12619927525520325, + "learning_rate": 2.3730969637464722e-06, + "loss": 0.0125, + "step": 99310 + }, + { + "epoch": 28.897294151876636, + "grad_norm": 0.11506721377372742, + "learning_rate": 2.370866087240158e-06, + "loss": 0.0105, + "step": 99320 + }, + { + "epoch": 28.90020366598778, + "grad_norm": 0.10955164581537247, + "learning_rate": 2.368636169833218e-06, + "loss": 0.0102, + "step": 99330 + }, + { + "epoch": 28.903113180098924, + "grad_norm": 0.08894036710262299, + "learning_rate": 2.3664072116950065e-06, + "loss": 0.01, + "step": 99340 + }, + { + "epoch": 28.906022694210066, + "grad_norm": 0.12434007227420807, + "learning_rate": 2.364179212994792e-06, + "loss": 0.0114, + "step": 99350 + }, + { + "epoch": 28.90893220832121, + "grad_norm": 0.10362933576107025, + "learning_rate": 2.361952173901773e-06, + "loss": 0.0114, + "step": 99360 + }, + { + "epoch": 28.911841722432353, + "grad_norm": 0.10673109441995621, + "learning_rate": 2.3597260945850842e-06, + "loss": 0.01, + "step": 99370 + }, + { + "epoch": 28.9147512365435, + "grad_norm": 0.12102537602186203, + "learning_rate": 2.3575009752137815e-06, + "loss": 0.0115, + "step": 99380 + }, + { + "epoch": 28.91766075065464, + "grad_norm": 0.35487326979637146, + "learning_rate": 2.3552768159568486e-06, + "loss": 0.0122, + "step": 99390 + }, + { + "epoch": 28.920570264765786, + "grad_norm": 0.09448620676994324, + "learning_rate": 2.3530536169831857e-06, + "loss": 0.0085, + "step": 99400 + }, + { + "epoch": 28.923479778876928, + "grad_norm": 0.12771224975585938, + "learning_rate": 2.3508313784616437e-06, + "loss": 0.0115, + "step": 99410 + }, + { + "epoch": 28.92638929298807, + "grad_norm": 0.095389224588871, + "learning_rate": 2.3486101005609828e-06, + "loss": 0.0126, + "step": 99420 + }, + { + "epoch": 28.929298807099215, + "grad_norm": 0.0789559856057167, + "learning_rate": 2.346389783449888e-06, + "loss": 0.0101, + "step": 99430 + }, + { + "epoch": 28.932208321210357, + "grad_norm": 0.09321089833974838, + "learning_rate": 2.344170427296985e-06, + "loss": 0.0103, + "step": 99440 + }, + { + "epoch": 28.935117835321503, + "grad_norm": 0.129488006234169, + "learning_rate": 2.3419520322708183e-06, + "loss": 0.0128, + "step": 99450 + }, + { + "epoch": 28.938027349432645, + "grad_norm": 0.07236841320991516, + "learning_rate": 2.33973459853986e-06, + "loss": 0.0121, + "step": 99460 + }, + { + "epoch": 28.940936863543786, + "grad_norm": 0.10917532444000244, + "learning_rate": 2.3375181262725036e-06, + "loss": 0.0122, + "step": 99470 + }, + { + "epoch": 28.943846377654932, + "grad_norm": 0.10473567247390747, + "learning_rate": 2.3353026156370816e-06, + "loss": 0.0093, + "step": 99480 + }, + { + "epoch": 28.946755891766074, + "grad_norm": 0.10416476428508759, + "learning_rate": 2.333088066801849e-06, + "loss": 0.0128, + "step": 99490 + }, + { + "epoch": 28.94966540587722, + "grad_norm": 0.1001383364200592, + "learning_rate": 2.33087447993498e-06, + "loss": 0.0118, + "step": 99500 + }, + { + "epoch": 28.95257491998836, + "grad_norm": 0.08039190620183945, + "learning_rate": 2.3286618552045856e-06, + "loss": 0.0092, + "step": 99510 + }, + { + "epoch": 28.955484434099507, + "grad_norm": 0.13723987340927124, + "learning_rate": 2.326450192778703e-06, + "loss": 0.0122, + "step": 99520 + }, + { + "epoch": 28.95839394821065, + "grad_norm": 0.11497245728969574, + "learning_rate": 2.3242394928252903e-06, + "loss": 0.0122, + "step": 99530 + }, + { + "epoch": 28.96130346232179, + "grad_norm": 0.15377052128314972, + "learning_rate": 2.3220297555122317e-06, + "loss": 0.0124, + "step": 99540 + }, + { + "epoch": 28.964212976432936, + "grad_norm": 0.14190872013568878, + "learning_rate": 2.3198209810073463e-06, + "loss": 0.0114, + "step": 99550 + }, + { + "epoch": 28.967122490544078, + "grad_norm": 0.101814866065979, + "learning_rate": 2.3176131694783782e-06, + "loss": 0.0104, + "step": 99560 + }, + { + "epoch": 28.970032004655224, + "grad_norm": 0.10409971326589584, + "learning_rate": 2.31540632109299e-06, + "loss": 0.0109, + "step": 99570 + }, + { + "epoch": 28.972941518766365, + "grad_norm": 0.142991840839386, + "learning_rate": 2.313200436018786e-06, + "loss": 0.0109, + "step": 99580 + }, + { + "epoch": 28.97585103287751, + "grad_norm": 0.12948636710643768, + "learning_rate": 2.3109955144232768e-06, + "loss": 0.0093, + "step": 99590 + }, + { + "epoch": 28.978760546988653, + "grad_norm": 0.09271764010190964, + "learning_rate": 2.308791556473921e-06, + "loss": 0.0104, + "step": 99600 + }, + { + "epoch": 28.981670061099795, + "grad_norm": 0.10465230792760849, + "learning_rate": 2.306588562338094e-06, + "loss": 0.0137, + "step": 99610 + }, + { + "epoch": 28.98457957521094, + "grad_norm": 0.14591771364212036, + "learning_rate": 2.304386532183092e-06, + "loss": 0.0142, + "step": 99620 + }, + { + "epoch": 28.987489089322082, + "grad_norm": 0.09760881960391998, + "learning_rate": 2.3021854661761534e-06, + "loss": 0.0125, + "step": 99630 + }, + { + "epoch": 28.990398603433228, + "grad_norm": 0.11471458524465561, + "learning_rate": 2.299985364484427e-06, + "loss": 0.0095, + "step": 99640 + }, + { + "epoch": 28.99330811754437, + "grad_norm": 0.09893053770065308, + "learning_rate": 2.297786227275003e-06, + "loss": 0.0102, + "step": 99650 + }, + { + "epoch": 28.996217631655515, + "grad_norm": 0.12702278792858124, + "learning_rate": 2.2955880547148833e-06, + "loss": 0.0111, + "step": 99660 + }, + { + "epoch": 28.999127145766657, + "grad_norm": 0.14084993302822113, + "learning_rate": 2.29339084697101e-06, + "loss": 0.0098, + "step": 99670 + }, + { + "epoch": 29.0020366598778, + "grad_norm": 0.09953160583972931, + "learning_rate": 2.291194604210249e-06, + "loss": 0.0111, + "step": 99680 + }, + { + "epoch": 29.004946173988944, + "grad_norm": 0.10567528754472733, + "learning_rate": 2.2889993265993875e-06, + "loss": 0.0088, + "step": 99690 + }, + { + "epoch": 29.007855688100086, + "grad_norm": 0.11675871163606644, + "learning_rate": 2.2868050143051343e-06, + "loss": 0.0114, + "step": 99700 + }, + { + "epoch": 29.010765202211232, + "grad_norm": 0.09852224588394165, + "learning_rate": 2.2846116674941465e-06, + "loss": 0.0115, + "step": 99710 + }, + { + "epoch": 29.013674716322374, + "grad_norm": 0.09203079342842102, + "learning_rate": 2.28241928633299e-06, + "loss": 0.0105, + "step": 99720 + }, + { + "epoch": 29.01658423043352, + "grad_norm": 0.12082663923501968, + "learning_rate": 2.2802278709881557e-06, + "loss": 0.0122, + "step": 99730 + }, + { + "epoch": 29.01949374454466, + "grad_norm": 0.10222533345222473, + "learning_rate": 2.27803742162607e-06, + "loss": 0.0091, + "step": 99740 + }, + { + "epoch": 29.022403258655803, + "grad_norm": 0.11994101852178574, + "learning_rate": 2.2758479384130903e-06, + "loss": 0.0106, + "step": 99750 + }, + { + "epoch": 29.02531277276695, + "grad_norm": 0.10856207460165024, + "learning_rate": 2.2736594215154865e-06, + "loss": 0.0092, + "step": 99760 + }, + { + "epoch": 29.02822228687809, + "grad_norm": 0.08752094954252243, + "learning_rate": 2.2714718710994594e-06, + "loss": 0.0115, + "step": 99770 + }, + { + "epoch": 29.031131800989236, + "grad_norm": 0.1148502454161644, + "learning_rate": 2.269285287331144e-06, + "loss": 0.0104, + "step": 99780 + }, + { + "epoch": 29.034041315100378, + "grad_norm": 0.11640161275863647, + "learning_rate": 2.2670996703765973e-06, + "loss": 0.0103, + "step": 99790 + }, + { + "epoch": 29.03695082921152, + "grad_norm": 0.18910947442054749, + "learning_rate": 2.264915020401799e-06, + "loss": 0.0111, + "step": 99800 + }, + { + "epoch": 29.039860343322665, + "grad_norm": 0.143405944108963, + "learning_rate": 2.2627313375726587e-06, + "loss": 0.0121, + "step": 99810 + }, + { + "epoch": 29.042769857433807, + "grad_norm": 0.10507558286190033, + "learning_rate": 2.2605486220550196e-06, + "loss": 0.0098, + "step": 99820 + }, + { + "epoch": 29.045679371544953, + "grad_norm": 0.249084934592247, + "learning_rate": 2.258366874014639e-06, + "loss": 0.0103, + "step": 99830 + }, + { + "epoch": 29.048588885656095, + "grad_norm": 0.09517933428287506, + "learning_rate": 2.256186093617204e-06, + "loss": 0.0128, + "step": 99840 + }, + { + "epoch": 29.05149839976724, + "grad_norm": 0.08562472462654114, + "learning_rate": 2.2540062810283326e-06, + "loss": 0.0106, + "step": 99850 + }, + { + "epoch": 29.054407913878382, + "grad_norm": 0.11253012716770172, + "learning_rate": 2.2518274364135716e-06, + "loss": 0.0115, + "step": 99860 + }, + { + "epoch": 29.057317427989524, + "grad_norm": 0.12640061974525452, + "learning_rate": 2.249649559938383e-06, + "loss": 0.01, + "step": 99870 + }, + { + "epoch": 29.06022694210067, + "grad_norm": 0.17042694985866547, + "learning_rate": 2.2474726517681683e-06, + "loss": 0.0095, + "step": 99880 + }, + { + "epoch": 29.06313645621181, + "grad_norm": 0.1118495836853981, + "learning_rate": 2.2452967120682444e-06, + "loss": 0.0096, + "step": 99890 + }, + { + "epoch": 29.066045970322957, + "grad_norm": 0.0998217910528183, + "learning_rate": 2.243121741003859e-06, + "loss": 0.0117, + "step": 99900 + }, + { + "epoch": 29.0689554844341, + "grad_norm": 0.08165398985147476, + "learning_rate": 2.240947738740195e-06, + "loss": 0.0096, + "step": 99910 + }, + { + "epoch": 29.071864998545244, + "grad_norm": 0.11531069874763489, + "learning_rate": 2.2387747054423436e-06, + "loss": 0.0116, + "step": 99920 + }, + { + "epoch": 29.074774512656386, + "grad_norm": 0.11632228642702103, + "learning_rate": 2.236602641275339e-06, + "loss": 0.0105, + "step": 99930 + }, + { + "epoch": 29.07768402676753, + "grad_norm": 0.13619601726531982, + "learning_rate": 2.2344315464041305e-06, + "loss": 0.0103, + "step": 99940 + }, + { + "epoch": 29.080593540878674, + "grad_norm": 0.13361766934394836, + "learning_rate": 2.232261420993603e-06, + "loss": 0.0118, + "step": 99950 + }, + { + "epoch": 29.083503054989816, + "grad_norm": 0.11400602012872696, + "learning_rate": 2.2300922652085592e-06, + "loss": 0.0106, + "step": 99960 + }, + { + "epoch": 29.08641256910096, + "grad_norm": 0.09594951570034027, + "learning_rate": 2.227924079213732e-06, + "loss": 0.0116, + "step": 99970 + }, + { + "epoch": 29.089322083212103, + "grad_norm": 0.12091630697250366, + "learning_rate": 2.2257568631737873e-06, + "loss": 0.0112, + "step": 99980 + }, + { + "epoch": 29.09223159732325, + "grad_norm": 0.11330015957355499, + "learning_rate": 2.223590617253306e-06, + "loss": 0.0118, + "step": 99990 + }, + { + "epoch": 29.09514111143439, + "grad_norm": 0.09932112693786621, + "learning_rate": 2.221425341616796e-06, + "loss": 0.0104, + "step": 100000 + }, + { + "epoch": 29.098050625545532, + "grad_norm": 0.15964160859584808, + "learning_rate": 2.2192610364286997e-06, + "loss": 0.0092, + "step": 100010 + }, + { + "epoch": 29.100960139656678, + "grad_norm": 0.11254847794771194, + "learning_rate": 2.2170977018533865e-06, + "loss": 0.0101, + "step": 100020 + }, + { + "epoch": 29.10386965376782, + "grad_norm": 0.1359476000070572, + "learning_rate": 2.214935338055141e-06, + "loss": 0.0106, + "step": 100030 + }, + { + "epoch": 29.106779167878965, + "grad_norm": 0.09331101924180984, + "learning_rate": 2.2127739451981803e-06, + "loss": 0.0103, + "step": 100040 + }, + { + "epoch": 29.109688681990107, + "grad_norm": 0.06974576413631439, + "learning_rate": 2.2106135234466545e-06, + "loss": 0.0101, + "step": 100050 + }, + { + "epoch": 29.112598196101253, + "grad_norm": 0.1043403372168541, + "learning_rate": 2.2084540729646286e-06, + "loss": 0.011, + "step": 100060 + }, + { + "epoch": 29.115507710212395, + "grad_norm": 0.27879592776298523, + "learning_rate": 2.206295593916096e-06, + "loss": 0.0103, + "step": 100070 + }, + { + "epoch": 29.118417224323537, + "grad_norm": 0.13096123933792114, + "learning_rate": 2.2041380864649814e-06, + "loss": 0.012, + "step": 100080 + }, + { + "epoch": 29.121326738434682, + "grad_norm": 0.09502067416906357, + "learning_rate": 2.2019815507751377e-06, + "loss": 0.0103, + "step": 100090 + }, + { + "epoch": 29.124236252545824, + "grad_norm": 0.12075458467006683, + "learning_rate": 2.1998259870103347e-06, + "loss": 0.0108, + "step": 100100 + }, + { + "epoch": 29.12714576665697, + "grad_norm": 0.09160366654396057, + "learning_rate": 2.1976713953342683e-06, + "loss": 0.0095, + "step": 100110 + }, + { + "epoch": 29.13005528076811, + "grad_norm": 0.09268689155578613, + "learning_rate": 2.1955177759105784e-06, + "loss": 0.0103, + "step": 100120 + }, + { + "epoch": 29.132964794879257, + "grad_norm": 0.12697972357273102, + "learning_rate": 2.193365128902811e-06, + "loss": 0.0099, + "step": 100130 + }, + { + "epoch": 29.1358743089904, + "grad_norm": 0.11594060808420181, + "learning_rate": 2.191213454474441e-06, + "loss": 0.011, + "step": 100140 + }, + { + "epoch": 29.13878382310154, + "grad_norm": 0.08746766299009323, + "learning_rate": 2.1890627527888807e-06, + "loss": 0.0112, + "step": 100150 + }, + { + "epoch": 29.141693337212686, + "grad_norm": 0.10066314786672592, + "learning_rate": 2.186913024009461e-06, + "loss": 0.0109, + "step": 100160 + }, + { + "epoch": 29.144602851323828, + "grad_norm": 0.09994985163211823, + "learning_rate": 2.184764268299436e-06, + "loss": 0.011, + "step": 100170 + }, + { + "epoch": 29.147512365434974, + "grad_norm": 0.11228153109550476, + "learning_rate": 2.1826164858219943e-06, + "loss": 0.0093, + "step": 100180 + }, + { + "epoch": 29.150421879546116, + "grad_norm": 0.08419906347990036, + "learning_rate": 2.1804696767402397e-06, + "loss": 0.0104, + "step": 100190 + }, + { + "epoch": 29.153331393657258, + "grad_norm": 0.16306298971176147, + "learning_rate": 2.1783238412172156e-06, + "loss": 0.013, + "step": 100200 + }, + { + "epoch": 29.156240907768403, + "grad_norm": 0.17592273652553558, + "learning_rate": 2.1761789794158763e-06, + "loss": 0.0142, + "step": 100210 + }, + { + "epoch": 29.159150421879545, + "grad_norm": 0.08122444897890091, + "learning_rate": 2.1740350914991135e-06, + "loss": 0.0098, + "step": 100220 + }, + { + "epoch": 29.16205993599069, + "grad_norm": 0.14341607689857483, + "learning_rate": 2.171892177629745e-06, + "loss": 0.0104, + "step": 100230 + }, + { + "epoch": 29.164969450101832, + "grad_norm": 0.08934833109378815, + "learning_rate": 2.1697502379705043e-06, + "loss": 0.0101, + "step": 100240 + }, + { + "epoch": 29.167878964212978, + "grad_norm": 0.10079644620418549, + "learning_rate": 2.167609272684064e-06, + "loss": 0.0103, + "step": 100250 + }, + { + "epoch": 29.17078847832412, + "grad_norm": 0.07793530821800232, + "learning_rate": 2.1654692819330095e-06, + "loss": 0.011, + "step": 100260 + }, + { + "epoch": 29.17369799243526, + "grad_norm": 0.11267593502998352, + "learning_rate": 2.1633302658798628e-06, + "loss": 0.0095, + "step": 100270 + }, + { + "epoch": 29.176607506546407, + "grad_norm": 0.07484762370586395, + "learning_rate": 2.16119222468707e-06, + "loss": 0.0109, + "step": 100280 + }, + { + "epoch": 29.17951702065755, + "grad_norm": 0.0997932106256485, + "learning_rate": 2.1590551585169992e-06, + "loss": 0.0094, + "step": 100290 + }, + { + "epoch": 29.182426534768695, + "grad_norm": 0.14049145579338074, + "learning_rate": 2.156919067531945e-06, + "loss": 0.0099, + "step": 100300 + }, + { + "epoch": 29.185336048879837, + "grad_norm": 0.11893592774868011, + "learning_rate": 2.154783951894129e-06, + "loss": 0.0107, + "step": 100310 + }, + { + "epoch": 29.188245562990982, + "grad_norm": 0.09182490408420563, + "learning_rate": 2.152649811765704e-06, + "loss": 0.0114, + "step": 100320 + }, + { + "epoch": 29.191155077102124, + "grad_norm": 0.0954454243183136, + "learning_rate": 2.150516647308738e-06, + "loss": 0.0094, + "step": 100330 + }, + { + "epoch": 29.194064591213266, + "grad_norm": 0.07581709325313568, + "learning_rate": 2.1483844586852325e-06, + "loss": 0.0099, + "step": 100340 + }, + { + "epoch": 29.19697410532441, + "grad_norm": 0.08594097942113876, + "learning_rate": 2.146253246057118e-06, + "loss": 0.0096, + "step": 100350 + }, + { + "epoch": 29.199883619435553, + "grad_norm": 0.06995853036642075, + "learning_rate": 2.1441230095862424e-06, + "loss": 0.0095, + "step": 100360 + }, + { + "epoch": 29.2027931335467, + "grad_norm": 0.10568541288375854, + "learning_rate": 2.141993749434379e-06, + "loss": 0.0089, + "step": 100370 + }, + { + "epoch": 29.20570264765784, + "grad_norm": 0.09679011255502701, + "learning_rate": 2.1398654657632347e-06, + "loss": 0.0095, + "step": 100380 + }, + { + "epoch": 29.208612161768986, + "grad_norm": 0.10506687313318253, + "learning_rate": 2.137738158734443e-06, + "loss": 0.0096, + "step": 100390 + }, + { + "epoch": 29.211521675880128, + "grad_norm": 0.0834222063422203, + "learning_rate": 2.135611828509554e-06, + "loss": 0.0111, + "step": 100400 + }, + { + "epoch": 29.21443118999127, + "grad_norm": 0.12911838293075562, + "learning_rate": 2.1334864752500434e-06, + "loss": 0.0114, + "step": 100410 + }, + { + "epoch": 29.217340704102416, + "grad_norm": 0.1603308469057083, + "learning_rate": 2.1313620991173288e-06, + "loss": 0.0117, + "step": 100420 + }, + { + "epoch": 29.220250218213558, + "grad_norm": 0.136185884475708, + "learning_rate": 2.1292387002727383e-06, + "loss": 0.0105, + "step": 100430 + }, + { + "epoch": 29.223159732324703, + "grad_norm": 0.07187611609697342, + "learning_rate": 2.127116278877524e-06, + "loss": 0.0107, + "step": 100440 + }, + { + "epoch": 29.226069246435845, + "grad_norm": 0.13732172548770905, + "learning_rate": 2.1249948350928755e-06, + "loss": 0.0114, + "step": 100450 + }, + { + "epoch": 29.22897876054699, + "grad_norm": 0.10140407830476761, + "learning_rate": 2.122874369079904e-06, + "loss": 0.0123, + "step": 100460 + }, + { + "epoch": 29.231888274658132, + "grad_norm": 0.13675770163536072, + "learning_rate": 2.120754880999644e-06, + "loss": 0.0114, + "step": 100470 + }, + { + "epoch": 29.234797788769274, + "grad_norm": 0.13266988098621368, + "learning_rate": 2.1186363710130518e-06, + "loss": 0.0121, + "step": 100480 + }, + { + "epoch": 29.23770730288042, + "grad_norm": 0.1064680740237236, + "learning_rate": 2.116518839281017e-06, + "loss": 0.01, + "step": 100490 + }, + { + "epoch": 29.24061681699156, + "grad_norm": 0.10683570057153702, + "learning_rate": 2.1144022859643557e-06, + "loss": 0.0102, + "step": 100500 + }, + { + "epoch": 29.243526331102707, + "grad_norm": 0.08171001076698303, + "learning_rate": 2.1122867112238016e-06, + "loss": 0.0107, + "step": 100510 + }, + { + "epoch": 29.24643584521385, + "grad_norm": 0.10762447118759155, + "learning_rate": 2.110172115220019e-06, + "loss": 0.0102, + "step": 100520 + }, + { + "epoch": 29.24934535932499, + "grad_norm": 0.10301747918128967, + "learning_rate": 2.108058498113601e-06, + "loss": 0.0109, + "step": 100530 + }, + { + "epoch": 29.252254873436137, + "grad_norm": 0.07178374379873276, + "learning_rate": 2.1059458600650593e-06, + "loss": 0.0118, + "step": 100540 + }, + { + "epoch": 29.25516438754728, + "grad_norm": 0.09938082098960876, + "learning_rate": 2.1038342012348394e-06, + "loss": 0.0109, + "step": 100550 + }, + { + "epoch": 29.258073901658424, + "grad_norm": 0.08361728489398956, + "learning_rate": 2.101723521783299e-06, + "loss": 0.0116, + "step": 100560 + }, + { + "epoch": 29.260983415769566, + "grad_norm": 0.12670157849788666, + "learning_rate": 2.099613821870742e-06, + "loss": 0.009, + "step": 100570 + }, + { + "epoch": 29.26389292988071, + "grad_norm": 0.11936090141534805, + "learning_rate": 2.0975051016573755e-06, + "loss": 0.0118, + "step": 100580 + }, + { + "epoch": 29.266802443991853, + "grad_norm": 0.09684336930513382, + "learning_rate": 2.0953973613033504e-06, + "loss": 0.0136, + "step": 100590 + }, + { + "epoch": 29.269711958102995, + "grad_norm": 0.0907241702079773, + "learning_rate": 2.0932906009687297e-06, + "loss": 0.0101, + "step": 100600 + }, + { + "epoch": 29.27262147221414, + "grad_norm": 0.09695664793252945, + "learning_rate": 2.0911848208135116e-06, + "loss": 0.0109, + "step": 100610 + }, + { + "epoch": 29.275530986325283, + "grad_norm": 0.10977193713188171, + "learning_rate": 2.0890800209976197e-06, + "loss": 0.0125, + "step": 100620 + }, + { + "epoch": 29.278440500436428, + "grad_norm": 0.09276364743709564, + "learning_rate": 2.0869762016808896e-06, + "loss": 0.0111, + "step": 100630 + }, + { + "epoch": 29.28135001454757, + "grad_norm": 0.2134445458650589, + "learning_rate": 2.0848733630231005e-06, + "loss": 0.0097, + "step": 100640 + }, + { + "epoch": 29.284259528658716, + "grad_norm": 0.0964401364326477, + "learning_rate": 2.082771505183949e-06, + "loss": 0.0085, + "step": 100650 + }, + { + "epoch": 29.287169042769857, + "grad_norm": 0.1553511768579483, + "learning_rate": 2.0806706283230563e-06, + "loss": 0.0111, + "step": 100660 + }, + { + "epoch": 29.290078556881, + "grad_norm": 0.09610084444284439, + "learning_rate": 2.0785707325999666e-06, + "loss": 0.0105, + "step": 100670 + }, + { + "epoch": 29.292988070992145, + "grad_norm": 0.08972111344337463, + "learning_rate": 2.076471818174155e-06, + "loss": 0.0107, + "step": 100680 + }, + { + "epoch": 29.295897585103287, + "grad_norm": 0.09561630338430405, + "learning_rate": 2.0743738852050243e-06, + "loss": 0.0104, + "step": 100690 + }, + { + "epoch": 29.298807099214432, + "grad_norm": 0.27085110545158386, + "learning_rate": 2.0722769338518955e-06, + "loss": 0.0099, + "step": 100700 + }, + { + "epoch": 29.301716613325574, + "grad_norm": 0.08971807360649109, + "learning_rate": 2.070180964274011e-06, + "loss": 0.0108, + "step": 100710 + }, + { + "epoch": 29.30462612743672, + "grad_norm": 0.12022259086370468, + "learning_rate": 2.068085976630559e-06, + "loss": 0.0108, + "step": 100720 + }, + { + "epoch": 29.30753564154786, + "grad_norm": 0.06886981427669525, + "learning_rate": 2.0659919710806346e-06, + "loss": 0.0099, + "step": 100730 + }, + { + "epoch": 29.310445155659004, + "grad_norm": 0.0815884917974472, + "learning_rate": 2.0638989477832598e-06, + "loss": 0.0099, + "step": 100740 + }, + { + "epoch": 29.31335466977015, + "grad_norm": 0.10361463576555252, + "learning_rate": 2.061806906897389e-06, + "loss": 0.0106, + "step": 100750 + }, + { + "epoch": 29.31626418388129, + "grad_norm": 0.11092673987150192, + "learning_rate": 2.0597158485819013e-06, + "loss": 0.0101, + "step": 100760 + }, + { + "epoch": 29.319173697992436, + "grad_norm": 0.11817178130149841, + "learning_rate": 2.057625772995598e-06, + "loss": 0.0121, + "step": 100770 + }, + { + "epoch": 29.32208321210358, + "grad_norm": 0.1493772715330124, + "learning_rate": 2.0555366802972015e-06, + "loss": 0.0115, + "step": 100780 + }, + { + "epoch": 29.324992726214724, + "grad_norm": 0.1793472021818161, + "learning_rate": 2.053448570645367e-06, + "loss": 0.0116, + "step": 100790 + }, + { + "epoch": 29.327902240325866, + "grad_norm": 0.11419256776571274, + "learning_rate": 2.0513614441986765e-06, + "loss": 0.0101, + "step": 100800 + }, + { + "epoch": 29.330811754437008, + "grad_norm": 0.092208132147789, + "learning_rate": 2.0492753011156303e-06, + "loss": 0.0104, + "step": 100810 + }, + { + "epoch": 29.333721268548153, + "grad_norm": 0.0990103930234909, + "learning_rate": 2.0471901415546556e-06, + "loss": 0.0103, + "step": 100820 + }, + { + "epoch": 29.336630782659295, + "grad_norm": 0.096556656062603, + "learning_rate": 2.0451059656741116e-06, + "loss": 0.0091, + "step": 100830 + }, + { + "epoch": 29.33954029677044, + "grad_norm": 0.13292334973812103, + "learning_rate": 2.043022773632275e-06, + "loss": 0.0104, + "step": 100840 + }, + { + "epoch": 29.342449810881583, + "grad_norm": 0.10213454067707062, + "learning_rate": 2.040940565587347e-06, + "loss": 0.0106, + "step": 100850 + }, + { + "epoch": 29.345359324992724, + "grad_norm": 0.09142401069402695, + "learning_rate": 2.0388593416974616e-06, + "loss": 0.0102, + "step": 100860 + }, + { + "epoch": 29.34826883910387, + "grad_norm": 0.14481885731220245, + "learning_rate": 2.0367791021206754e-06, + "loss": 0.0109, + "step": 100870 + }, + { + "epoch": 29.351178353215012, + "grad_norm": 0.1479727029800415, + "learning_rate": 2.0346998470149626e-06, + "loss": 0.012, + "step": 100880 + }, + { + "epoch": 29.354087867326157, + "grad_norm": 0.12052427977323532, + "learning_rate": 2.032621576538238e-06, + "loss": 0.0097, + "step": 100890 + }, + { + "epoch": 29.3569973814373, + "grad_norm": 0.14890700578689575, + "learning_rate": 2.030544290848322e-06, + "loss": 0.0116, + "step": 100900 + }, + { + "epoch": 29.359906895548445, + "grad_norm": 0.12221146374940872, + "learning_rate": 2.0284679901029766e-06, + "loss": 0.0112, + "step": 100910 + }, + { + "epoch": 29.362816409659587, + "grad_norm": 0.09359100461006165, + "learning_rate": 2.026392674459888e-06, + "loss": 0.0106, + "step": 100920 + }, + { + "epoch": 29.36572592377073, + "grad_norm": 0.10174845159053802, + "learning_rate": 2.0243183440766517e-06, + "loss": 0.0095, + "step": 100930 + }, + { + "epoch": 29.368635437881874, + "grad_norm": 0.10905132442712784, + "learning_rate": 2.0222449991108094e-06, + "loss": 0.0121, + "step": 100940 + }, + { + "epoch": 29.371544951993016, + "grad_norm": 0.10710779577493668, + "learning_rate": 2.0201726397198104e-06, + "loss": 0.0103, + "step": 100950 + }, + { + "epoch": 29.37445446610416, + "grad_norm": 0.13362687826156616, + "learning_rate": 2.0181012660610443e-06, + "loss": 0.0134, + "step": 100960 + }, + { + "epoch": 29.377363980215303, + "grad_norm": 0.0569881796836853, + "learning_rate": 2.0160308782918095e-06, + "loss": 0.0107, + "step": 100970 + }, + { + "epoch": 29.38027349432645, + "grad_norm": 0.12120698392391205, + "learning_rate": 2.0139614765693426e-06, + "loss": 0.0103, + "step": 100980 + }, + { + "epoch": 29.38318300843759, + "grad_norm": 0.11228722333908081, + "learning_rate": 2.0118930610508058e-06, + "loss": 0.0122, + "step": 100990 + }, + { + "epoch": 29.386092522548733, + "grad_norm": 0.15557563304901123, + "learning_rate": 2.009825631893275e-06, + "loss": 0.0106, + "step": 101000 + }, + { + "epoch": 29.38900203665988, + "grad_norm": 0.14742694795131683, + "learning_rate": 2.007759189253755e-06, + "loss": 0.0119, + "step": 101010 + }, + { + "epoch": 29.39191155077102, + "grad_norm": 0.09245942533016205, + "learning_rate": 2.0056937332891882e-06, + "loss": 0.0099, + "step": 101020 + }, + { + "epoch": 29.394821064882166, + "grad_norm": 0.17287573218345642, + "learning_rate": 2.003629264156428e-06, + "loss": 0.0099, + "step": 101030 + }, + { + "epoch": 29.397730578993308, + "grad_norm": 0.10807964950799942, + "learning_rate": 2.0015657820122534e-06, + "loss": 0.0094, + "step": 101040 + }, + { + "epoch": 29.400640093104453, + "grad_norm": 0.13151443004608154, + "learning_rate": 1.999503287013376e-06, + "loss": 0.01, + "step": 101050 + }, + { + "epoch": 29.403549607215595, + "grad_norm": 0.1426835060119629, + "learning_rate": 1.9974417793164305e-06, + "loss": 0.0103, + "step": 101060 + }, + { + "epoch": 29.406459121326737, + "grad_norm": 0.0971461832523346, + "learning_rate": 1.9953812590779735e-06, + "loss": 0.0129, + "step": 101070 + }, + { + "epoch": 29.409368635437882, + "grad_norm": 0.08606117218732834, + "learning_rate": 1.993321726454483e-06, + "loss": 0.0104, + "step": 101080 + }, + { + "epoch": 29.412278149549024, + "grad_norm": 0.17918094992637634, + "learning_rate": 1.9912631816023718e-06, + "loss": 0.0105, + "step": 101090 + }, + { + "epoch": 29.41518766366017, + "grad_norm": 0.0989542081952095, + "learning_rate": 1.9892056246779737e-06, + "loss": 0.0112, + "step": 101100 + }, + { + "epoch": 29.418097177771312, + "grad_norm": 0.07978363335132599, + "learning_rate": 1.9871490558375404e-06, + "loss": 0.0104, + "step": 101110 + }, + { + "epoch": 29.421006691882457, + "grad_norm": 0.10640457272529602, + "learning_rate": 1.985093475237262e-06, + "loss": 0.01, + "step": 101120 + }, + { + "epoch": 29.4239162059936, + "grad_norm": 0.07783504575490952, + "learning_rate": 1.983038883033245e-06, + "loss": 0.0099, + "step": 101130 + }, + { + "epoch": 29.42682572010474, + "grad_norm": 0.11495430767536163, + "learning_rate": 1.980985279381519e-06, + "loss": 0.0093, + "step": 101140 + }, + { + "epoch": 29.429735234215887, + "grad_norm": 0.12741905450820923, + "learning_rate": 1.9789326644380407e-06, + "loss": 0.0092, + "step": 101150 + }, + { + "epoch": 29.43264474832703, + "grad_norm": 0.11289866268634796, + "learning_rate": 1.976881038358696e-06, + "loss": 0.0104, + "step": 101160 + }, + { + "epoch": 29.435554262438174, + "grad_norm": 0.08221495896577835, + "learning_rate": 1.9748304012992925e-06, + "loss": 0.0107, + "step": 101170 + }, + { + "epoch": 29.438463776549316, + "grad_norm": 0.07474654912948608, + "learning_rate": 1.9727807534155602e-06, + "loss": 0.0108, + "step": 101180 + }, + { + "epoch": 29.441373290660458, + "grad_norm": 0.09814850986003876, + "learning_rate": 1.97073209486316e-06, + "loss": 0.0112, + "step": 101190 + }, + { + "epoch": 29.444282804771603, + "grad_norm": 0.10816729813814163, + "learning_rate": 1.9686844257976683e-06, + "loss": 0.0104, + "step": 101200 + }, + { + "epoch": 29.447192318882745, + "grad_norm": 0.09502719342708588, + "learning_rate": 1.966637746374598e-06, + "loss": 0.0125, + "step": 101210 + }, + { + "epoch": 29.45010183299389, + "grad_norm": 0.09641389548778534, + "learning_rate": 1.9645920567493753e-06, + "loss": 0.0103, + "step": 101220 + }, + { + "epoch": 29.453011347105033, + "grad_norm": 0.09855595231056213, + "learning_rate": 1.96254735707736e-06, + "loss": 0.0094, + "step": 101230 + }, + { + "epoch": 29.45592086121618, + "grad_norm": 0.06335611641407013, + "learning_rate": 1.9605036475138368e-06, + "loss": 0.0112, + "step": 101240 + }, + { + "epoch": 29.45883037532732, + "grad_norm": 0.1334671974182129, + "learning_rate": 1.9584609282140043e-06, + "loss": 0.0112, + "step": 101250 + }, + { + "epoch": 29.461739889438462, + "grad_norm": 0.11383145302534103, + "learning_rate": 1.9564191993330003e-06, + "loss": 0.0116, + "step": 101260 + }, + { + "epoch": 29.464649403549608, + "grad_norm": 0.14884626865386963, + "learning_rate": 1.9543784610258764e-06, + "loss": 0.011, + "step": 101270 + }, + { + "epoch": 29.46755891766075, + "grad_norm": 0.10305126011371613, + "learning_rate": 1.952338713447613e-06, + "loss": 0.0109, + "step": 101280 + }, + { + "epoch": 29.470468431771895, + "grad_norm": 0.1396833211183548, + "learning_rate": 1.950299956753121e-06, + "loss": 0.0123, + "step": 101290 + }, + { + "epoch": 29.473377945883037, + "grad_norm": 0.09662694483995438, + "learning_rate": 1.948262191097227e-06, + "loss": 0.0115, + "step": 101300 + }, + { + "epoch": 29.476287459994182, + "grad_norm": 0.10248537361621857, + "learning_rate": 1.946225416634684e-06, + "loss": 0.0112, + "step": 101310 + }, + { + "epoch": 29.479196974105324, + "grad_norm": 0.08456873893737793, + "learning_rate": 1.944189633520171e-06, + "loss": 0.0115, + "step": 101320 + }, + { + "epoch": 29.482106488216466, + "grad_norm": 0.07734515517950058, + "learning_rate": 1.9421548419082978e-06, + "loss": 0.0101, + "step": 101330 + }, + { + "epoch": 29.48501600232761, + "grad_norm": 0.1411316692829132, + "learning_rate": 1.9401210419535883e-06, + "loss": 0.0104, + "step": 101340 + }, + { + "epoch": 29.487925516438754, + "grad_norm": 0.111363984644413, + "learning_rate": 1.938088233810496e-06, + "loss": 0.0091, + "step": 101350 + }, + { + "epoch": 29.4908350305499, + "grad_norm": 0.07467706501483917, + "learning_rate": 1.9360564176334068e-06, + "loss": 0.0104, + "step": 101360 + }, + { + "epoch": 29.49374454466104, + "grad_norm": 0.0759497657418251, + "learning_rate": 1.9340255935766164e-06, + "loss": 0.0106, + "step": 101370 + }, + { + "epoch": 29.496654058772187, + "grad_norm": 0.1001879870891571, + "learning_rate": 1.9319957617943505e-06, + "loss": 0.0093, + "step": 101380 + }, + { + "epoch": 29.49956357288333, + "grad_norm": 0.11414728313684464, + "learning_rate": 1.9299669224407674e-06, + "loss": 0.0078, + "step": 101390 + }, + { + "epoch": 29.50247308699447, + "grad_norm": 0.11445130407810211, + "learning_rate": 1.9279390756699435e-06, + "loss": 0.012, + "step": 101400 + }, + { + "epoch": 29.505382601105616, + "grad_norm": 0.1151626780629158, + "learning_rate": 1.925912221635879e-06, + "loss": 0.009, + "step": 101410 + }, + { + "epoch": 29.508292115216758, + "grad_norm": 0.10042371600866318, + "learning_rate": 1.9238863604924937e-06, + "loss": 0.0103, + "step": 101420 + }, + { + "epoch": 29.511201629327903, + "grad_norm": 0.11989793181419373, + "learning_rate": 1.9218614923936508e-06, + "loss": 0.0094, + "step": 101430 + }, + { + "epoch": 29.514111143439045, + "grad_norm": 0.15080603957176208, + "learning_rate": 1.919837617493119e-06, + "loss": 0.0108, + "step": 101440 + }, + { + "epoch": 29.51702065755019, + "grad_norm": 0.08879242092370987, + "learning_rate": 1.9178147359445953e-06, + "loss": 0.0117, + "step": 101450 + }, + { + "epoch": 29.519930171661333, + "grad_norm": 0.14672519266605377, + "learning_rate": 1.915792847901707e-06, + "loss": 0.0108, + "step": 101460 + }, + { + "epoch": 29.522839685772475, + "grad_norm": 0.12680429220199585, + "learning_rate": 1.9137719535180077e-06, + "loss": 0.0113, + "step": 101470 + }, + { + "epoch": 29.52574919988362, + "grad_norm": 0.11752855777740479, + "learning_rate": 1.9117520529469652e-06, + "loss": 0.0104, + "step": 101480 + }, + { + "epoch": 29.528658713994762, + "grad_norm": 0.12188045680522919, + "learning_rate": 1.9097331463419805e-06, + "loss": 0.0105, + "step": 101490 + }, + { + "epoch": 29.531568228105908, + "grad_norm": 0.05813829228281975, + "learning_rate": 1.9077152338563724e-06, + "loss": 0.0104, + "step": 101500 + }, + { + "epoch": 29.53447774221705, + "grad_norm": 0.09863369166851044, + "learning_rate": 1.9056983156433939e-06, + "loss": 0.0099, + "step": 101510 + }, + { + "epoch": 29.537387256328195, + "grad_norm": 0.12428983300924301, + "learning_rate": 1.9036823918562102e-06, + "loss": 0.0125, + "step": 101520 + }, + { + "epoch": 29.540296770439337, + "grad_norm": 0.1044844314455986, + "learning_rate": 1.9016674626479208e-06, + "loss": 0.0117, + "step": 101530 + }, + { + "epoch": 29.54320628455048, + "grad_norm": 0.12420742958784103, + "learning_rate": 1.8996535281715483e-06, + "loss": 0.0105, + "step": 101540 + }, + { + "epoch": 29.546115798661624, + "grad_norm": 0.13989852368831635, + "learning_rate": 1.8976405885800335e-06, + "loss": 0.0107, + "step": 101550 + }, + { + "epoch": 29.549025312772766, + "grad_norm": 0.10063424706459045, + "learning_rate": 1.895628644026251e-06, + "loss": 0.0106, + "step": 101560 + }, + { + "epoch": 29.55193482688391, + "grad_norm": 0.08218695968389511, + "learning_rate": 1.8936176946629886e-06, + "loss": 0.0098, + "step": 101570 + }, + { + "epoch": 29.554844340995054, + "grad_norm": 0.1434507519006729, + "learning_rate": 1.8916077406429706e-06, + "loss": 0.0124, + "step": 101580 + }, + { + "epoch": 29.557753855106196, + "grad_norm": 0.10283982753753662, + "learning_rate": 1.8895987821188348e-06, + "loss": 0.01, + "step": 101590 + }, + { + "epoch": 29.56066336921734, + "grad_norm": 0.07965666800737381, + "learning_rate": 1.8875908192431512e-06, + "loss": 0.0101, + "step": 101600 + }, + { + "epoch": 29.563572883328483, + "grad_norm": 0.13167135417461395, + "learning_rate": 1.8855838521684093e-06, + "loss": 0.01, + "step": 101610 + }, + { + "epoch": 29.56648239743963, + "grad_norm": 0.110337033867836, + "learning_rate": 1.8835778810470255e-06, + "loss": 0.0112, + "step": 101620 + }, + { + "epoch": 29.56939191155077, + "grad_norm": 0.09886447340250015, + "learning_rate": 1.8815729060313448e-06, + "loss": 0.0099, + "step": 101630 + }, + { + "epoch": 29.572301425661916, + "grad_norm": 0.11939137428998947, + "learning_rate": 1.8795689272736232e-06, + "loss": 0.0108, + "step": 101640 + }, + { + "epoch": 29.575210939773058, + "grad_norm": 0.08539621531963348, + "learning_rate": 1.8775659449260545e-06, + "loss": 0.0129, + "step": 101650 + }, + { + "epoch": 29.5781204538842, + "grad_norm": 0.10148588567972183, + "learning_rate": 1.8755639591407554e-06, + "loss": 0.0093, + "step": 101660 + }, + { + "epoch": 29.581029967995345, + "grad_norm": 0.11697600036859512, + "learning_rate": 1.8735629700697604e-06, + "loss": 0.0097, + "step": 101670 + }, + { + "epoch": 29.583939482106487, + "grad_norm": 0.08920827507972717, + "learning_rate": 1.8715629778650267e-06, + "loss": 0.0103, + "step": 101680 + }, + { + "epoch": 29.586848996217633, + "grad_norm": 0.13706074655056, + "learning_rate": 1.8695639826784461e-06, + "loss": 0.0123, + "step": 101690 + }, + { + "epoch": 29.589758510328775, + "grad_norm": 0.15870434045791626, + "learning_rate": 1.8675659846618287e-06, + "loss": 0.0112, + "step": 101700 + }, + { + "epoch": 29.59266802443992, + "grad_norm": 0.08806022256612778, + "learning_rate": 1.86556898396691e-06, + "loss": 0.0094, + "step": 101710 + }, + { + "epoch": 29.595577538551062, + "grad_norm": 0.11320655792951584, + "learning_rate": 1.86357298074534e-06, + "loss": 0.0129, + "step": 101720 + }, + { + "epoch": 29.598487052662204, + "grad_norm": 0.28692761063575745, + "learning_rate": 1.861577975148716e-06, + "loss": 0.0089, + "step": 101730 + }, + { + "epoch": 29.60139656677335, + "grad_norm": 0.11090050637722015, + "learning_rate": 1.8595839673285386e-06, + "loss": 0.0118, + "step": 101740 + }, + { + "epoch": 29.60430608088449, + "grad_norm": 0.13300001621246338, + "learning_rate": 1.8575909574362365e-06, + "loss": 0.0115, + "step": 101750 + }, + { + "epoch": 29.607215594995637, + "grad_norm": 0.07972249388694763, + "learning_rate": 1.8555989456231687e-06, + "loss": 0.0123, + "step": 101760 + }, + { + "epoch": 29.61012510910678, + "grad_norm": 0.14499983191490173, + "learning_rate": 1.8536079320406191e-06, + "loss": 0.0111, + "step": 101770 + }, + { + "epoch": 29.613034623217924, + "grad_norm": 0.11848033219575882, + "learning_rate": 1.8516179168397885e-06, + "loss": 0.0102, + "step": 101780 + }, + { + "epoch": 29.615944137329066, + "grad_norm": 0.10531267523765564, + "learning_rate": 1.8496289001718025e-06, + "loss": 0.0103, + "step": 101790 + }, + { + "epoch": 29.618853651440208, + "grad_norm": 0.08717234432697296, + "learning_rate": 1.8476408821877156e-06, + "loss": 0.0131, + "step": 101800 + }, + { + "epoch": 29.621763165551354, + "grad_norm": 0.12764336168766022, + "learning_rate": 1.8456538630385084e-06, + "loss": 0.0112, + "step": 101810 + }, + { + "epoch": 29.624672679662496, + "grad_norm": 0.14781703054904938, + "learning_rate": 1.843667842875077e-06, + "loss": 0.011, + "step": 101820 + }, + { + "epoch": 29.62758219377364, + "grad_norm": 0.12605774402618408, + "learning_rate": 1.841682821848249e-06, + "loss": 0.0114, + "step": 101830 + }, + { + "epoch": 29.630491707884783, + "grad_norm": 0.09296493977308273, + "learning_rate": 1.8396988001087756e-06, + "loss": 0.0126, + "step": 101840 + }, + { + "epoch": 29.633401221995925, + "grad_norm": 0.10579824447631836, + "learning_rate": 1.8377157778073261e-06, + "loss": 0.0096, + "step": 101850 + }, + { + "epoch": 29.63631073610707, + "grad_norm": 0.10037817060947418, + "learning_rate": 1.835733755094502e-06, + "loss": 0.0095, + "step": 101860 + }, + { + "epoch": 29.639220250218212, + "grad_norm": 0.07390552014112473, + "learning_rate": 1.8337527321208193e-06, + "loss": 0.0138, + "step": 101870 + }, + { + "epoch": 29.642129764329358, + "grad_norm": 0.1700415462255478, + "learning_rate": 1.8317727090367292e-06, + "loss": 0.0091, + "step": 101880 + }, + { + "epoch": 29.6450392784405, + "grad_norm": 0.10085412859916687, + "learning_rate": 1.8297936859925984e-06, + "loss": 0.0108, + "step": 101890 + }, + { + "epoch": 29.647948792551645, + "grad_norm": 0.11723766475915909, + "learning_rate": 1.8278156631387216e-06, + "loss": 0.012, + "step": 101900 + }, + { + "epoch": 29.650858306662787, + "grad_norm": 0.1000712588429451, + "learning_rate": 1.8258386406253152e-06, + "loss": 0.0095, + "step": 101910 + }, + { + "epoch": 29.65376782077393, + "grad_norm": 0.12283973395824432, + "learning_rate": 1.8238626186025226e-06, + "loss": 0.0112, + "step": 101920 + }, + { + "epoch": 29.656677334885075, + "grad_norm": 0.10673637688159943, + "learning_rate": 1.8218875972204101e-06, + "loss": 0.0097, + "step": 101930 + }, + { + "epoch": 29.659586848996216, + "grad_norm": 0.12367621809244156, + "learning_rate": 1.8199135766289648e-06, + "loss": 0.0103, + "step": 101940 + }, + { + "epoch": 29.662496363107362, + "grad_norm": 0.08036265522241592, + "learning_rate": 1.8179405569781032e-06, + "loss": 0.0105, + "step": 101950 + }, + { + "epoch": 29.665405877218504, + "grad_norm": 0.118735671043396, + "learning_rate": 1.8159685384176638e-06, + "loss": 0.0104, + "step": 101960 + }, + { + "epoch": 29.66831539132965, + "grad_norm": 0.1949208378791809, + "learning_rate": 1.813997521097407e-06, + "loss": 0.0119, + "step": 101970 + }, + { + "epoch": 29.67122490544079, + "grad_norm": 0.12386653572320938, + "learning_rate": 1.812027505167016e-06, + "loss": 0.0114, + "step": 101980 + }, + { + "epoch": 29.674134419551933, + "grad_norm": 0.1726614385843277, + "learning_rate": 1.810058490776103e-06, + "loss": 0.0114, + "step": 101990 + }, + { + "epoch": 29.67704393366308, + "grad_norm": 0.08835361897945404, + "learning_rate": 1.8080904780742052e-06, + "loss": 0.0106, + "step": 102000 + }, + { + "epoch": 29.67995344777422, + "grad_norm": 0.09935850650072098, + "learning_rate": 1.8061234672107747e-06, + "loss": 0.0119, + "step": 102010 + }, + { + "epoch": 29.682862961885366, + "grad_norm": 0.1313389390707016, + "learning_rate": 1.80415745833519e-06, + "loss": 0.0099, + "step": 102020 + }, + { + "epoch": 29.685772475996508, + "grad_norm": 0.15757817029953003, + "learning_rate": 1.802192451596767e-06, + "loss": 0.0112, + "step": 102030 + }, + { + "epoch": 29.688681990107654, + "grad_norm": 0.08893778175115585, + "learning_rate": 1.8002284471447311e-06, + "loss": 0.0143, + "step": 102040 + }, + { + "epoch": 29.691591504218795, + "grad_norm": 0.0861179456114769, + "learning_rate": 1.7982654451282282e-06, + "loss": 0.0095, + "step": 102050 + }, + { + "epoch": 29.694501018329937, + "grad_norm": 0.0965023934841156, + "learning_rate": 1.7963034456963424e-06, + "loss": 0.0103, + "step": 102060 + }, + { + "epoch": 29.697410532441083, + "grad_norm": 0.1216684877872467, + "learning_rate": 1.7943424489980758e-06, + "loss": 0.0128, + "step": 102070 + }, + { + "epoch": 29.700320046552225, + "grad_norm": 0.09902364760637283, + "learning_rate": 1.7923824551823499e-06, + "loss": 0.01, + "step": 102080 + }, + { + "epoch": 29.70322956066337, + "grad_norm": 0.11828017234802246, + "learning_rate": 1.7904234643980117e-06, + "loss": 0.0116, + "step": 102090 + }, + { + "epoch": 29.706139074774512, + "grad_norm": 0.11265537142753601, + "learning_rate": 1.788465476793834e-06, + "loss": 0.0108, + "step": 102100 + }, + { + "epoch": 29.709048588885658, + "grad_norm": 0.08566176146268845, + "learning_rate": 1.78650849251852e-06, + "loss": 0.0121, + "step": 102110 + }, + { + "epoch": 29.7119581029968, + "grad_norm": 0.27605873346328735, + "learning_rate": 1.7845525117206802e-06, + "loss": 0.0114, + "step": 102120 + }, + { + "epoch": 29.71486761710794, + "grad_norm": 0.07160203158855438, + "learning_rate": 1.7825975345488626e-06, + "loss": 0.0102, + "step": 102130 + }, + { + "epoch": 29.717777131219087, + "grad_norm": 0.11537927389144897, + "learning_rate": 1.7806435611515387e-06, + "loss": 0.0109, + "step": 102140 + }, + { + "epoch": 29.72068664533023, + "grad_norm": 0.18365629017353058, + "learning_rate": 1.7786905916770962e-06, + "loss": 0.01, + "step": 102150 + }, + { + "epoch": 29.723596159441374, + "grad_norm": 0.11191767454147339, + "learning_rate": 1.7767386262738483e-06, + "loss": 0.0104, + "step": 102160 + }, + { + "epoch": 29.726505673552516, + "grad_norm": 0.1435288041830063, + "learning_rate": 1.774787665090033e-06, + "loss": 0.0098, + "step": 102170 + }, + { + "epoch": 29.729415187663662, + "grad_norm": 0.09715932607650757, + "learning_rate": 1.7728377082738218e-06, + "loss": 0.0106, + "step": 102180 + }, + { + "epoch": 29.732324701774804, + "grad_norm": 0.15426063537597656, + "learning_rate": 1.7708887559732894e-06, + "loss": 0.0099, + "step": 102190 + }, + { + "epoch": 29.735234215885946, + "grad_norm": 0.09964562952518463, + "learning_rate": 1.768940808336456e-06, + "loss": 0.0099, + "step": 102200 + }, + { + "epoch": 29.73814372999709, + "grad_norm": 0.12124449759721756, + "learning_rate": 1.7669938655112467e-06, + "loss": 0.0105, + "step": 102210 + }, + { + "epoch": 29.741053244108233, + "grad_norm": 0.1147916316986084, + "learning_rate": 1.7650479276455244e-06, + "loss": 0.0099, + "step": 102220 + }, + { + "epoch": 29.74396275821938, + "grad_norm": 0.14069874584674835, + "learning_rate": 1.7631029948870713e-06, + "loss": 0.0118, + "step": 102230 + }, + { + "epoch": 29.74687227233052, + "grad_norm": 0.12165278196334839, + "learning_rate": 1.7611590673835859e-06, + "loss": 0.0082, + "step": 102240 + }, + { + "epoch": 29.749781786441666, + "grad_norm": 0.09283885359764099, + "learning_rate": 1.759216145282705e-06, + "loss": 0.0115, + "step": 102250 + }, + { + "epoch": 29.752691300552808, + "grad_norm": 0.06405047327280045, + "learning_rate": 1.757274228731972e-06, + "loss": 0.0097, + "step": 102260 + }, + { + "epoch": 29.75560081466395, + "grad_norm": 0.14725351333618164, + "learning_rate": 1.755333317878871e-06, + "loss": 0.0096, + "step": 102270 + }, + { + "epoch": 29.758510328775095, + "grad_norm": 0.12912625074386597, + "learning_rate": 1.753393412870794e-06, + "loss": 0.0114, + "step": 102280 + }, + { + "epoch": 29.761419842886237, + "grad_norm": 0.07354626804590225, + "learning_rate": 1.751454513855068e-06, + "loss": 0.0104, + "step": 102290 + }, + { + "epoch": 29.764329356997383, + "grad_norm": 0.10961515456438065, + "learning_rate": 1.7495166209789404e-06, + "loss": 0.0106, + "step": 102300 + }, + { + "epoch": 29.767238871108525, + "grad_norm": 0.08846455812454224, + "learning_rate": 1.74757973438958e-06, + "loss": 0.0095, + "step": 102310 + }, + { + "epoch": 29.770148385219667, + "grad_norm": 0.09661451727151871, + "learning_rate": 1.7456438542340758e-06, + "loss": 0.0097, + "step": 102320 + }, + { + "epoch": 29.773057899330812, + "grad_norm": 0.20122584700584412, + "learning_rate": 1.7437089806594553e-06, + "loss": 0.0103, + "step": 102330 + }, + { + "epoch": 29.775967413441954, + "grad_norm": 0.050072576850652695, + "learning_rate": 1.7417751138126525e-06, + "loss": 0.0103, + "step": 102340 + }, + { + "epoch": 29.7788769275531, + "grad_norm": 0.10244257003068924, + "learning_rate": 1.7398422538405317e-06, + "loss": 0.0084, + "step": 102350 + }, + { + "epoch": 29.78178644166424, + "grad_norm": 0.1100853830575943, + "learning_rate": 1.7379104008898805e-06, + "loss": 0.0107, + "step": 102360 + }, + { + "epoch": 29.784695955775387, + "grad_norm": 0.1796657145023346, + "learning_rate": 1.7359795551074165e-06, + "loss": 0.01, + "step": 102370 + }, + { + "epoch": 29.78760546988653, + "grad_norm": 0.12397977709770203, + "learning_rate": 1.7340497166397707e-06, + "loss": 0.0113, + "step": 102380 + }, + { + "epoch": 29.79051498399767, + "grad_norm": 0.08723917603492737, + "learning_rate": 1.7321208856334974e-06, + "loss": 0.0112, + "step": 102390 + }, + { + "epoch": 29.793424498108816, + "grad_norm": 0.14286616444587708, + "learning_rate": 1.7301930622350832e-06, + "loss": 0.0091, + "step": 102400 + }, + { + "epoch": 29.79633401221996, + "grad_norm": 0.06366188824176788, + "learning_rate": 1.7282662465909355e-06, + "loss": 0.0092, + "step": 102410 + }, + { + "epoch": 29.799243526331104, + "grad_norm": 0.13555972278118134, + "learning_rate": 1.7263404388473807e-06, + "loss": 0.0109, + "step": 102420 + }, + { + "epoch": 29.802153040442246, + "grad_norm": 0.23167648911476135, + "learning_rate": 1.7244156391506654e-06, + "loss": 0.0101, + "step": 102430 + }, + { + "epoch": 29.80506255455339, + "grad_norm": 0.0960502177476883, + "learning_rate": 1.7224918476469758e-06, + "loss": 0.0101, + "step": 102440 + }, + { + "epoch": 29.807972068664533, + "grad_norm": 0.1119316816329956, + "learning_rate": 1.7205690644824068e-06, + "loss": 0.0123, + "step": 102450 + }, + { + "epoch": 29.810881582775675, + "grad_norm": 0.12297827750444412, + "learning_rate": 1.7186472898029782e-06, + "loss": 0.0102, + "step": 102460 + }, + { + "epoch": 29.81379109688682, + "grad_norm": 0.12361941486597061, + "learning_rate": 1.7167265237546382e-06, + "loss": 0.0128, + "step": 102470 + }, + { + "epoch": 29.816700610997962, + "grad_norm": 0.11140082031488419, + "learning_rate": 1.71480676648326e-06, + "loss": 0.0117, + "step": 102480 + }, + { + "epoch": 29.819610125109108, + "grad_norm": 0.11519839614629745, + "learning_rate": 1.7128880181346307e-06, + "loss": 0.0151, + "step": 102490 + }, + { + "epoch": 29.82251963922025, + "grad_norm": 0.16340799629688263, + "learning_rate": 1.7109702788544718e-06, + "loss": 0.0115, + "step": 102500 + }, + { + "epoch": 29.825429153331395, + "grad_norm": 0.13667264580726624, + "learning_rate": 1.7090535487884167e-06, + "loss": 0.0103, + "step": 102510 + }, + { + "epoch": 29.828338667442537, + "grad_norm": 0.0913395956158638, + "learning_rate": 1.707137828082036e-06, + "loss": 0.012, + "step": 102520 + }, + { + "epoch": 29.83124818155368, + "grad_norm": 0.15245436131954193, + "learning_rate": 1.705223116880808e-06, + "loss": 0.0102, + "step": 102530 + }, + { + "epoch": 29.834157695664825, + "grad_norm": 0.1046397015452385, + "learning_rate": 1.7033094153301466e-06, + "loss": 0.011, + "step": 102540 + }, + { + "epoch": 29.837067209775967, + "grad_norm": 0.08179458230733871, + "learning_rate": 1.7013967235753886e-06, + "loss": 0.0113, + "step": 102550 + }, + { + "epoch": 29.839976723887112, + "grad_norm": 0.13974329829216003, + "learning_rate": 1.699485041761783e-06, + "loss": 0.0101, + "step": 102560 + }, + { + "epoch": 29.842886237998254, + "grad_norm": 0.1355580985546112, + "learning_rate": 1.6975743700345154e-06, + "loss": 0.0099, + "step": 102570 + }, + { + "epoch": 29.845795752109396, + "grad_norm": 0.08333700150251389, + "learning_rate": 1.695664708538683e-06, + "loss": 0.0115, + "step": 102580 + }, + { + "epoch": 29.84870526622054, + "grad_norm": 0.16347995400428772, + "learning_rate": 1.6937560574193145e-06, + "loss": 0.0107, + "step": 102590 + }, + { + "epoch": 29.851614780331683, + "grad_norm": 0.07840604335069656, + "learning_rate": 1.6918484168213627e-06, + "loss": 0.0126, + "step": 102600 + }, + { + "epoch": 29.85452429444283, + "grad_norm": 0.09157735109329224, + "learning_rate": 1.6899417868896983e-06, + "loss": 0.0128, + "step": 102610 + }, + { + "epoch": 29.85743380855397, + "grad_norm": 0.1562497913837433, + "learning_rate": 1.688036167769112e-06, + "loss": 0.0111, + "step": 102620 + }, + { + "epoch": 29.860343322665116, + "grad_norm": 0.09241738170385361, + "learning_rate": 1.6861315596043297e-06, + "loss": 0.0111, + "step": 102630 + }, + { + "epoch": 29.863252836776258, + "grad_norm": 0.08529426902532578, + "learning_rate": 1.6842279625399925e-06, + "loss": 0.012, + "step": 102640 + }, + { + "epoch": 29.8661623508874, + "grad_norm": 0.11504929512739182, + "learning_rate": 1.6823253767206632e-06, + "loss": 0.0106, + "step": 102650 + }, + { + "epoch": 29.869071864998546, + "grad_norm": 0.14017443358898163, + "learning_rate": 1.6804238022908326e-06, + "loss": 0.0132, + "step": 102660 + }, + { + "epoch": 29.871981379109688, + "grad_norm": 0.10535626113414764, + "learning_rate": 1.6785232393949152e-06, + "loss": 0.0095, + "step": 102670 + }, + { + "epoch": 29.874890893220833, + "grad_norm": 0.09237831085920334, + "learning_rate": 1.6766236881772439e-06, + "loss": 0.0107, + "step": 102680 + }, + { + "epoch": 29.877800407331975, + "grad_norm": 0.1218201294541359, + "learning_rate": 1.6747251487820736e-06, + "loss": 0.0087, + "step": 102690 + }, + { + "epoch": 29.88070992144312, + "grad_norm": 0.18164633214473724, + "learning_rate": 1.6728276213535903e-06, + "loss": 0.0131, + "step": 102700 + }, + { + "epoch": 29.883619435554262, + "grad_norm": 0.09517299383878708, + "learning_rate": 1.6709311060358989e-06, + "loss": 0.01, + "step": 102710 + }, + { + "epoch": 29.886528949665404, + "grad_norm": 0.06624863296747208, + "learning_rate": 1.6690356029730274e-06, + "loss": 0.0096, + "step": 102720 + }, + { + "epoch": 29.88943846377655, + "grad_norm": 0.09960609674453735, + "learning_rate": 1.6671411123089175e-06, + "loss": 0.0117, + "step": 102730 + }, + { + "epoch": 29.89234797788769, + "grad_norm": 0.12041423469781876, + "learning_rate": 1.6652476341874572e-06, + "loss": 0.0115, + "step": 102740 + }, + { + "epoch": 29.895257491998837, + "grad_norm": 0.12794938683509827, + "learning_rate": 1.6633551687524385e-06, + "loss": 0.0106, + "step": 102750 + }, + { + "epoch": 29.89816700610998, + "grad_norm": 0.250029593706131, + "learning_rate": 1.661463716147576e-06, + "loss": 0.011, + "step": 102760 + }, + { + "epoch": 29.901076520221125, + "grad_norm": 0.1258418709039688, + "learning_rate": 1.6595732765165184e-06, + "loss": 0.0135, + "step": 102770 + }, + { + "epoch": 29.903986034332267, + "grad_norm": 0.06732446700334549, + "learning_rate": 1.657683850002834e-06, + "loss": 0.0095, + "step": 102780 + }, + { + "epoch": 29.90689554844341, + "grad_norm": 0.2555878758430481, + "learning_rate": 1.65579543675001e-06, + "loss": 0.0108, + "step": 102790 + }, + { + "epoch": 29.909805062554554, + "grad_norm": 0.09978805482387543, + "learning_rate": 1.6539080369014564e-06, + "loss": 0.0121, + "step": 102800 + }, + { + "epoch": 29.912714576665696, + "grad_norm": 0.07008997350931168, + "learning_rate": 1.6520216506005087e-06, + "loss": 0.0087, + "step": 102810 + }, + { + "epoch": 29.91562409077684, + "grad_norm": 0.09952642023563385, + "learning_rate": 1.650136277990432e-06, + "loss": 0.0101, + "step": 102820 + }, + { + "epoch": 29.918533604887983, + "grad_norm": 0.10367520153522491, + "learning_rate": 1.6482519192144007e-06, + "loss": 0.0084, + "step": 102830 + }, + { + "epoch": 29.92144311899913, + "grad_norm": 0.0878778025507927, + "learning_rate": 1.6463685744155215e-06, + "loss": 0.0103, + "step": 102840 + }, + { + "epoch": 29.92435263311027, + "grad_norm": 0.13769274950027466, + "learning_rate": 1.6444862437368268e-06, + "loss": 0.0103, + "step": 102850 + }, + { + "epoch": 29.927262147221413, + "grad_norm": 0.10060884803533554, + "learning_rate": 1.642604927321259e-06, + "loss": 0.0105, + "step": 102860 + }, + { + "epoch": 29.930171661332558, + "grad_norm": 0.12200016528367996, + "learning_rate": 1.6407246253116987e-06, + "loss": 0.0111, + "step": 102870 + }, + { + "epoch": 29.9330811754437, + "grad_norm": 0.09440630674362183, + "learning_rate": 1.6388453378509383e-06, + "loss": 0.0103, + "step": 102880 + }, + { + "epoch": 29.935990689554846, + "grad_norm": 0.10546430200338364, + "learning_rate": 1.6369670650817004e-06, + "loss": 0.0085, + "step": 102890 + }, + { + "epoch": 29.938900203665987, + "grad_norm": 0.12097833305597305, + "learning_rate": 1.6350898071466225e-06, + "loss": 0.0092, + "step": 102900 + }, + { + "epoch": 29.941809717777133, + "grad_norm": 0.10115401446819305, + "learning_rate": 1.6332135641882757e-06, + "loss": 0.0113, + "step": 102910 + }, + { + "epoch": 29.944719231888275, + "grad_norm": 0.09432811290025711, + "learning_rate": 1.6313383363491424e-06, + "loss": 0.0121, + "step": 102920 + }, + { + "epoch": 29.947628745999417, + "grad_norm": 0.10614863783121109, + "learning_rate": 1.6294641237716373e-06, + "loss": 0.0114, + "step": 102930 + }, + { + "epoch": 29.950538260110562, + "grad_norm": 0.10308197885751724, + "learning_rate": 1.6275909265980982e-06, + "loss": 0.0096, + "step": 102940 + }, + { + "epoch": 29.953447774221704, + "grad_norm": 0.1290087252855301, + "learning_rate": 1.6257187449707745e-06, + "loss": 0.0091, + "step": 102950 + }, + { + "epoch": 29.95635728833285, + "grad_norm": 0.13921836018562317, + "learning_rate": 1.6238475790318492e-06, + "loss": 0.0119, + "step": 102960 + }, + { + "epoch": 29.95926680244399, + "grad_norm": 0.13044220209121704, + "learning_rate": 1.6219774289234285e-06, + "loss": 0.0112, + "step": 102970 + }, + { + "epoch": 29.962176316555134, + "grad_norm": 0.09974444657564163, + "learning_rate": 1.620108294787534e-06, + "loss": 0.0103, + "step": 102980 + }, + { + "epoch": 29.96508583066628, + "grad_norm": 0.12263932824134827, + "learning_rate": 1.618240176766112e-06, + "loss": 0.009, + "step": 102990 + }, + { + "epoch": 29.96799534477742, + "grad_norm": 0.09024327248334885, + "learning_rate": 1.6163730750010376e-06, + "loss": 0.0107, + "step": 103000 + }, + { + "epoch": 29.970904858888566, + "grad_norm": 0.11445499956607819, + "learning_rate": 1.6145069896341069e-06, + "loss": 0.0101, + "step": 103010 + }, + { + "epoch": 29.97381437299971, + "grad_norm": 0.1532578021287918, + "learning_rate": 1.6126419208070336e-06, + "loss": 0.0119, + "step": 103020 + }, + { + "epoch": 29.976723887110854, + "grad_norm": 0.12690888345241547, + "learning_rate": 1.610777868661451e-06, + "loss": 0.0134, + "step": 103030 + }, + { + "epoch": 29.979633401221996, + "grad_norm": 0.14670616388320923, + "learning_rate": 1.6089148333389342e-06, + "loss": 0.0097, + "step": 103040 + }, + { + "epoch": 29.982542915333138, + "grad_norm": 0.07866153120994568, + "learning_rate": 1.6070528149809632e-06, + "loss": 0.0103, + "step": 103050 + }, + { + "epoch": 29.985452429444283, + "grad_norm": 0.083701491355896, + "learning_rate": 1.6051918137289417e-06, + "loss": 0.011, + "step": 103060 + }, + { + "epoch": 29.988361943555425, + "grad_norm": 0.1177658662199974, + "learning_rate": 1.6033318297242033e-06, + "loss": 0.0139, + "step": 103070 + }, + { + "epoch": 29.99127145766657, + "grad_norm": 0.08951467275619507, + "learning_rate": 1.601472863108005e-06, + "loss": 0.0101, + "step": 103080 + }, + { + "epoch": 29.994180971777713, + "grad_norm": 0.0805337056517601, + "learning_rate": 1.5996149140215206e-06, + "loss": 0.0109, + "step": 103090 + }, + { + "epoch": 29.997090485888858, + "grad_norm": 0.09919805079698563, + "learning_rate": 1.597757982605847e-06, + "loss": 0.0106, + "step": 103100 + }, + { + "epoch": 30.0, + "grad_norm": 0.08726580440998077, + "learning_rate": 1.595902069002005e-06, + "loss": 0.0109, + "step": 103110 + }, + { + "epoch": 30.002909514111142, + "grad_norm": 0.08814594149589539, + "learning_rate": 1.5940471733509465e-06, + "loss": 0.0092, + "step": 103120 + }, + { + "epoch": 30.005819028222287, + "grad_norm": 0.12121989578008652, + "learning_rate": 1.5921932957935292e-06, + "loss": 0.0103, + "step": 103130 + }, + { + "epoch": 30.00872854233343, + "grad_norm": 0.07614349573850632, + "learning_rate": 1.5903404364705488e-06, + "loss": 0.0108, + "step": 103140 + }, + { + "epoch": 30.011638056444575, + "grad_norm": 0.07821208983659744, + "learning_rate": 1.5884885955227173e-06, + "loss": 0.011, + "step": 103150 + }, + { + "epoch": 30.014547570555717, + "grad_norm": 0.1182958260178566, + "learning_rate": 1.5866377730906678e-06, + "loss": 0.0111, + "step": 103160 + }, + { + "epoch": 30.017457084666862, + "grad_norm": 0.07547006011009216, + "learning_rate": 1.5847879693149609e-06, + "loss": 0.0097, + "step": 103170 + }, + { + "epoch": 30.020366598778004, + "grad_norm": 0.1237194836139679, + "learning_rate": 1.5829391843360707e-06, + "loss": 0.0102, + "step": 103180 + }, + { + "epoch": 30.023276112889146, + "grad_norm": 0.3577120304107666, + "learning_rate": 1.5810914182944085e-06, + "loss": 0.0118, + "step": 103190 + }, + { + "epoch": 30.02618562700029, + "grad_norm": 0.19005011022090912, + "learning_rate": 1.5792446713302937e-06, + "loss": 0.0106, + "step": 103200 + }, + { + "epoch": 30.029095141111434, + "grad_norm": 0.0836147665977478, + "learning_rate": 1.5773989435839792e-06, + "loss": 0.0098, + "step": 103210 + }, + { + "epoch": 30.03200465522258, + "grad_norm": 0.08408933877944946, + "learning_rate": 1.5755542351956308e-06, + "loss": 0.01, + "step": 103220 + }, + { + "epoch": 30.03491416933372, + "grad_norm": 0.09453455358743668, + "learning_rate": 1.5737105463053452e-06, + "loss": 0.0113, + "step": 103230 + }, + { + "epoch": 30.037823683444866, + "grad_norm": 0.0822778269648552, + "learning_rate": 1.5718678770531403e-06, + "loss": 0.0098, + "step": 103240 + }, + { + "epoch": 30.04073319755601, + "grad_norm": 0.1148940920829773, + "learning_rate": 1.5700262275789489e-06, + "loss": 0.0115, + "step": 103250 + }, + { + "epoch": 30.04364271166715, + "grad_norm": 0.10709510743618011, + "learning_rate": 1.5681855980226377e-06, + "loss": 0.0114, + "step": 103260 + }, + { + "epoch": 30.046552225778296, + "grad_norm": 0.13144202530384064, + "learning_rate": 1.5663459885239868e-06, + "loss": 0.0108, + "step": 103270 + }, + { + "epoch": 30.049461739889438, + "grad_norm": 0.09734860807657242, + "learning_rate": 1.5645073992227056e-06, + "loss": 0.0111, + "step": 103280 + }, + { + "epoch": 30.052371254000583, + "grad_norm": 0.09503549337387085, + "learning_rate": 1.5626698302584174e-06, + "loss": 0.0112, + "step": 103290 + }, + { + "epoch": 30.055280768111725, + "grad_norm": 0.11342082172632217, + "learning_rate": 1.5608332817706777e-06, + "loss": 0.0082, + "step": 103300 + }, + { + "epoch": 30.058190282222867, + "grad_norm": 0.10683822631835938, + "learning_rate": 1.558997753898963e-06, + "loss": 0.0114, + "step": 103310 + }, + { + "epoch": 30.061099796334013, + "grad_norm": 0.21133284270763397, + "learning_rate": 1.5571632467826652e-06, + "loss": 0.009, + "step": 103320 + }, + { + "epoch": 30.064009310445154, + "grad_norm": 0.18397383391857147, + "learning_rate": 1.5553297605610978e-06, + "loss": 0.0118, + "step": 103330 + }, + { + "epoch": 30.0669188245563, + "grad_norm": 0.11281268298625946, + "learning_rate": 1.5534972953735143e-06, + "loss": 0.0128, + "step": 103340 + }, + { + "epoch": 30.069828338667442, + "grad_norm": 0.12641794979572296, + "learning_rate": 1.551665851359072e-06, + "loss": 0.0118, + "step": 103350 + }, + { + "epoch": 30.072737852778587, + "grad_norm": 0.09797398746013641, + "learning_rate": 1.5498354286568545e-06, + "loss": 0.01, + "step": 103360 + }, + { + "epoch": 30.07564736688973, + "grad_norm": 0.12295550107955933, + "learning_rate": 1.5480060274058721e-06, + "loss": 0.0112, + "step": 103370 + }, + { + "epoch": 30.07855688100087, + "grad_norm": 0.09329306334257126, + "learning_rate": 1.546177647745059e-06, + "loss": 0.0099, + "step": 103380 + }, + { + "epoch": 30.081466395112017, + "grad_norm": 0.13914833962917328, + "learning_rate": 1.5443502898132672e-06, + "loss": 0.0094, + "step": 103390 + }, + { + "epoch": 30.08437590922316, + "grad_norm": 0.08172451704740524, + "learning_rate": 1.5425239537492674e-06, + "loss": 0.0106, + "step": 103400 + }, + { + "epoch": 30.087285423334304, + "grad_norm": 0.17728687822818756, + "learning_rate": 1.5406986396917622e-06, + "loss": 0.0111, + "step": 103410 + }, + { + "epoch": 30.090194937445446, + "grad_norm": 0.15300533175468445, + "learning_rate": 1.5388743477793737e-06, + "loss": 0.0112, + "step": 103420 + }, + { + "epoch": 30.09310445155659, + "grad_norm": 0.1140894889831543, + "learning_rate": 1.5370510781506415e-06, + "loss": 0.0101, + "step": 103430 + }, + { + "epoch": 30.096013965667733, + "grad_norm": 0.14600563049316406, + "learning_rate": 1.535228830944031e-06, + "loss": 0.0103, + "step": 103440 + }, + { + "epoch": 30.098923479778875, + "grad_norm": 0.09308206290006638, + "learning_rate": 1.5334076062979335e-06, + "loss": 0.0111, + "step": 103450 + }, + { + "epoch": 30.10183299389002, + "grad_norm": 0.1037123054265976, + "learning_rate": 1.5315874043506584e-06, + "loss": 0.0115, + "step": 103460 + }, + { + "epoch": 30.104742508001163, + "grad_norm": 0.09141115099191666, + "learning_rate": 1.5297682252404316e-06, + "loss": 0.0113, + "step": 103470 + }, + { + "epoch": 30.10765202211231, + "grad_norm": 0.14962445199489594, + "learning_rate": 1.5279500691054127e-06, + "loss": 0.0099, + "step": 103480 + }, + { + "epoch": 30.11056153622345, + "grad_norm": 0.08548431098461151, + "learning_rate": 1.5261329360836813e-06, + "loss": 0.0094, + "step": 103490 + }, + { + "epoch": 30.113471050334596, + "grad_norm": 0.10140138864517212, + "learning_rate": 1.52431682631323e-06, + "loss": 0.012, + "step": 103500 + }, + { + "epoch": 30.116380564445738, + "grad_norm": 0.1146552711725235, + "learning_rate": 1.522501739931989e-06, + "loss": 0.0109, + "step": 103510 + }, + { + "epoch": 30.11929007855688, + "grad_norm": 0.10760214924812317, + "learning_rate": 1.5206876770777928e-06, + "loss": 0.0098, + "step": 103520 + }, + { + "epoch": 30.122199592668025, + "grad_norm": 0.07790767401456833, + "learning_rate": 1.5188746378884111e-06, + "loss": 0.0104, + "step": 103530 + }, + { + "epoch": 30.125109106779167, + "grad_norm": 0.10486675798892975, + "learning_rate": 1.517062622501537e-06, + "loss": 0.0107, + "step": 103540 + }, + { + "epoch": 30.128018620890312, + "grad_norm": 0.11626142263412476, + "learning_rate": 1.5152516310547727e-06, + "loss": 0.0105, + "step": 103550 + }, + { + "epoch": 30.130928135001454, + "grad_norm": 0.12500493228435516, + "learning_rate": 1.5134416636856607e-06, + "loss": 0.0105, + "step": 103560 + }, + { + "epoch": 30.1338376491126, + "grad_norm": 0.05703049153089523, + "learning_rate": 1.5116327205316465e-06, + "loss": 0.0098, + "step": 103570 + }, + { + "epoch": 30.136747163223742, + "grad_norm": 0.09867025911808014, + "learning_rate": 1.5098248017301136e-06, + "loss": 0.0118, + "step": 103580 + }, + { + "epoch": 30.139656677334884, + "grad_norm": 0.14735692739486694, + "learning_rate": 1.5080179074183587e-06, + "loss": 0.0106, + "step": 103590 + }, + { + "epoch": 30.14256619144603, + "grad_norm": 0.07283208519220352, + "learning_rate": 1.5062120377336036e-06, + "loss": 0.0094, + "step": 103600 + }, + { + "epoch": 30.14547570555717, + "grad_norm": 0.06981806457042694, + "learning_rate": 1.5044071928129953e-06, + "loss": 0.011, + "step": 103610 + }, + { + "epoch": 30.148385219668317, + "grad_norm": 0.13346993923187256, + "learning_rate": 1.5026033727935978e-06, + "loss": 0.0094, + "step": 103620 + }, + { + "epoch": 30.15129473377946, + "grad_norm": 0.09835748374462128, + "learning_rate": 1.5008005778123963e-06, + "loss": 0.012, + "step": 103630 + }, + { + "epoch": 30.154204247890604, + "grad_norm": 0.13341672718524933, + "learning_rate": 1.4989988080063043e-06, + "loss": 0.0111, + "step": 103640 + }, + { + "epoch": 30.157113762001746, + "grad_norm": 0.1406274288892746, + "learning_rate": 1.4971980635121562e-06, + "loss": 0.0115, + "step": 103650 + }, + { + "epoch": 30.160023276112888, + "grad_norm": 0.11412177234888077, + "learning_rate": 1.4953983444667025e-06, + "loss": 0.0078, + "step": 103660 + }, + { + "epoch": 30.162932790224033, + "grad_norm": 0.12772245705127716, + "learning_rate": 1.4935996510066203e-06, + "loss": 0.0132, + "step": 103670 + }, + { + "epoch": 30.165842304335175, + "grad_norm": 0.11617162823677063, + "learning_rate": 1.4918019832685136e-06, + "loss": 0.0135, + "step": 103680 + }, + { + "epoch": 30.16875181844632, + "grad_norm": 0.13743850588798523, + "learning_rate": 1.490005341388902e-06, + "loss": 0.0108, + "step": 103690 + }, + { + "epoch": 30.171661332557463, + "grad_norm": 0.0694182813167572, + "learning_rate": 1.488209725504221e-06, + "loss": 0.0098, + "step": 103700 + }, + { + "epoch": 30.174570846668605, + "grad_norm": 0.08494656533002853, + "learning_rate": 1.4864151357508433e-06, + "loss": 0.0106, + "step": 103710 + }, + { + "epoch": 30.17748036077975, + "grad_norm": 0.15668514370918274, + "learning_rate": 1.4846215722650547e-06, + "loss": 0.0119, + "step": 103720 + }, + { + "epoch": 30.180389874890892, + "grad_norm": 0.08049424737691879, + "learning_rate": 1.4828290351830665e-06, + "loss": 0.0099, + "step": 103730 + }, + { + "epoch": 30.183299389002038, + "grad_norm": 0.12102431803941727, + "learning_rate": 1.4810375246409996e-06, + "loss": 0.0117, + "step": 103740 + }, + { + "epoch": 30.18620890311318, + "grad_norm": 0.09686911851167679, + "learning_rate": 1.479247040774922e-06, + "loss": 0.0099, + "step": 103750 + }, + { + "epoch": 30.189118417224325, + "grad_norm": 0.07911285758018494, + "learning_rate": 1.4774575837208032e-06, + "loss": 0.0106, + "step": 103760 + }, + { + "epoch": 30.192027931335467, + "grad_norm": 0.24151480197906494, + "learning_rate": 1.475669153614536e-06, + "loss": 0.0105, + "step": 103770 + }, + { + "epoch": 30.19493744544661, + "grad_norm": 0.09205321967601776, + "learning_rate": 1.473881750591944e-06, + "loss": 0.0092, + "step": 103780 + }, + { + "epoch": 30.197846959557754, + "grad_norm": 0.11365154385566711, + "learning_rate": 1.472095374788771e-06, + "loss": 0.011, + "step": 103790 + }, + { + "epoch": 30.200756473668896, + "grad_norm": 0.11532478034496307, + "learning_rate": 1.470310026340676e-06, + "loss": 0.0126, + "step": 103800 + }, + { + "epoch": 30.20366598778004, + "grad_norm": 0.09694591909646988, + "learning_rate": 1.4685257053832485e-06, + "loss": 0.0102, + "step": 103810 + }, + { + "epoch": 30.206575501891184, + "grad_norm": 0.08770394325256348, + "learning_rate": 1.46674241205199e-06, + "loss": 0.01, + "step": 103820 + }, + { + "epoch": 30.20948501600233, + "grad_norm": 0.2188090980052948, + "learning_rate": 1.4649601464823376e-06, + "loss": 0.0102, + "step": 103830 + }, + { + "epoch": 30.21239453011347, + "grad_norm": 0.09261366724967957, + "learning_rate": 1.463178908809636e-06, + "loss": 0.0112, + "step": 103840 + }, + { + "epoch": 30.215304044224613, + "grad_norm": 0.08838114887475967, + "learning_rate": 1.4613986991691603e-06, + "loss": 0.0108, + "step": 103850 + }, + { + "epoch": 30.21821355833576, + "grad_norm": 0.11517535895109177, + "learning_rate": 1.4596195176961092e-06, + "loss": 0.0131, + "step": 103860 + }, + { + "epoch": 30.2211230724469, + "grad_norm": 0.09030836820602417, + "learning_rate": 1.4578413645255945e-06, + "loss": 0.0098, + "step": 103870 + }, + { + "epoch": 30.224032586558046, + "grad_norm": 0.08262136578559875, + "learning_rate": 1.4560642397926599e-06, + "loss": 0.0103, + "step": 103880 + }, + { + "epoch": 30.226942100669188, + "grad_norm": 0.11506786942481995, + "learning_rate": 1.4542881436322625e-06, + "loss": 0.0111, + "step": 103890 + }, + { + "epoch": 30.229851614780333, + "grad_norm": 0.09458038955926895, + "learning_rate": 1.452513076179286e-06, + "loss": 0.0096, + "step": 103900 + }, + { + "epoch": 30.232761128891475, + "grad_norm": 0.08717934042215347, + "learning_rate": 1.4507390375685376e-06, + "loss": 0.0096, + "step": 103910 + }, + { + "epoch": 30.235670643002617, + "grad_norm": 0.12714262306690216, + "learning_rate": 1.4489660279347427e-06, + "loss": 0.0106, + "step": 103920 + }, + { + "epoch": 30.238580157113763, + "grad_norm": 0.08668507635593414, + "learning_rate": 1.4471940474125472e-06, + "loss": 0.0111, + "step": 103930 + }, + { + "epoch": 30.241489671224905, + "grad_norm": 0.09792734682559967, + "learning_rate": 1.4454230961365217e-06, + "loss": 0.0122, + "step": 103940 + }, + { + "epoch": 30.24439918533605, + "grad_norm": 0.18155668675899506, + "learning_rate": 1.4436531742411617e-06, + "loss": 0.0129, + "step": 103950 + }, + { + "epoch": 30.247308699447192, + "grad_norm": 0.11382405459880829, + "learning_rate": 1.4418842818608785e-06, + "loss": 0.0107, + "step": 103960 + }, + { + "epoch": 30.250218213558334, + "grad_norm": 0.08610507100820541, + "learning_rate": 1.4401164191300076e-06, + "loss": 0.0103, + "step": 103970 + }, + { + "epoch": 30.25312772766948, + "grad_norm": 0.127118319272995, + "learning_rate": 1.4383495861828083e-06, + "loss": 0.0104, + "step": 103980 + }, + { + "epoch": 30.25603724178062, + "grad_norm": 0.06659460067749023, + "learning_rate": 1.43658378315346e-06, + "loss": 0.0094, + "step": 103990 + }, + { + "epoch": 30.258946755891767, + "grad_norm": 0.13005906343460083, + "learning_rate": 1.4348190101760606e-06, + "loss": 0.0098, + "step": 104000 + }, + { + "epoch": 30.26185627000291, + "grad_norm": 0.0763126015663147, + "learning_rate": 1.4330552673846342e-06, + "loss": 0.0087, + "step": 104010 + }, + { + "epoch": 30.264765784114054, + "grad_norm": 0.11550036072731018, + "learning_rate": 1.4312925549131289e-06, + "loss": 0.0105, + "step": 104020 + }, + { + "epoch": 30.267675298225196, + "grad_norm": 0.1373172104358673, + "learning_rate": 1.4295308728954093e-06, + "loss": 0.011, + "step": 104030 + }, + { + "epoch": 30.270584812336338, + "grad_norm": 0.09596370160579681, + "learning_rate": 1.4277702214652566e-06, + "loss": 0.0104, + "step": 104040 + }, + { + "epoch": 30.273494326447484, + "grad_norm": 0.08254432678222656, + "learning_rate": 1.4260106007563906e-06, + "loss": 0.0093, + "step": 104050 + }, + { + "epoch": 30.276403840558626, + "grad_norm": 0.11744855344295502, + "learning_rate": 1.424252010902441e-06, + "loss": 0.0097, + "step": 104060 + }, + { + "epoch": 30.27931335466977, + "grad_norm": 0.11121942102909088, + "learning_rate": 1.422494452036956e-06, + "loss": 0.009, + "step": 104070 + }, + { + "epoch": 30.282222868780913, + "grad_norm": 0.07434450089931488, + "learning_rate": 1.4207379242934121e-06, + "loss": 0.0103, + "step": 104080 + }, + { + "epoch": 30.28513238289206, + "grad_norm": 0.11534038931131363, + "learning_rate": 1.418982427805211e-06, + "loss": 0.0101, + "step": 104090 + }, + { + "epoch": 30.2880418970032, + "grad_norm": 0.10008417814970016, + "learning_rate": 1.4172279627056695e-06, + "loss": 0.0102, + "step": 104100 + }, + { + "epoch": 30.290951411114342, + "grad_norm": 0.11260826885700226, + "learning_rate": 1.4154745291280209e-06, + "loss": 0.0106, + "step": 104110 + }, + { + "epoch": 30.293860925225488, + "grad_norm": 0.10742580890655518, + "learning_rate": 1.4137221272054318e-06, + "loss": 0.0103, + "step": 104120 + }, + { + "epoch": 30.29677043933663, + "grad_norm": 0.0617227703332901, + "learning_rate": 1.4119707570709895e-06, + "loss": 0.0105, + "step": 104130 + }, + { + "epoch": 30.299679953447775, + "grad_norm": 0.06543669104576111, + "learning_rate": 1.4102204188576905e-06, + "loss": 0.0125, + "step": 104140 + }, + { + "epoch": 30.302589467558917, + "grad_norm": 0.13847072422504425, + "learning_rate": 1.4084711126984674e-06, + "loss": 0.0105, + "step": 104150 + }, + { + "epoch": 30.305498981670063, + "grad_norm": 0.09959860146045685, + "learning_rate": 1.4067228387261682e-06, + "loss": 0.0106, + "step": 104160 + }, + { + "epoch": 30.308408495781205, + "grad_norm": 0.09151283651590347, + "learning_rate": 1.4049755970735606e-06, + "loss": 0.0097, + "step": 104170 + }, + { + "epoch": 30.311318009892346, + "grad_norm": 0.0935925841331482, + "learning_rate": 1.4032293878733398e-06, + "loss": 0.0112, + "step": 104180 + }, + { + "epoch": 30.314227524003492, + "grad_norm": 0.10972609370946884, + "learning_rate": 1.4014842112581117e-06, + "loss": 0.0105, + "step": 104190 + }, + { + "epoch": 30.317137038114634, + "grad_norm": 0.11533210426568985, + "learning_rate": 1.39974006736042e-06, + "loss": 0.0102, + "step": 104200 + }, + { + "epoch": 30.32004655222578, + "grad_norm": 0.15704567730426788, + "learning_rate": 1.397996956312712e-06, + "loss": 0.0119, + "step": 104210 + }, + { + "epoch": 30.32295606633692, + "grad_norm": 0.10478635877370834, + "learning_rate": 1.3962548782473723e-06, + "loss": 0.0097, + "step": 104220 + }, + { + "epoch": 30.325865580448067, + "grad_norm": 0.10519915819168091, + "learning_rate": 1.3945138332966962e-06, + "loss": 0.0103, + "step": 104230 + }, + { + "epoch": 30.32877509455921, + "grad_norm": 0.05585957318544388, + "learning_rate": 1.3927738215929064e-06, + "loss": 0.0094, + "step": 104240 + }, + { + "epoch": 30.33168460867035, + "grad_norm": 0.1267285794019699, + "learning_rate": 1.3910348432681458e-06, + "loss": 0.0101, + "step": 104250 + }, + { + "epoch": 30.334594122781496, + "grad_norm": 0.08199922740459442, + "learning_rate": 1.3892968984544767e-06, + "loss": 0.0104, + "step": 104260 + }, + { + "epoch": 30.337503636892638, + "grad_norm": 0.22397685050964355, + "learning_rate": 1.387559987283884e-06, + "loss": 0.0115, + "step": 104270 + }, + { + "epoch": 30.340413151003784, + "grad_norm": 0.07445159554481506, + "learning_rate": 1.38582410988828e-06, + "loss": 0.0097, + "step": 104280 + }, + { + "epoch": 30.343322665114925, + "grad_norm": 0.0937233492732048, + "learning_rate": 1.384089266399488e-06, + "loss": 0.0106, + "step": 104290 + }, + { + "epoch": 30.34623217922607, + "grad_norm": 0.10164239257574081, + "learning_rate": 1.3823554569492557e-06, + "loss": 0.009, + "step": 104300 + }, + { + "epoch": 30.349141693337213, + "grad_norm": 0.1221945658326149, + "learning_rate": 1.3806226816692595e-06, + "loss": 0.011, + "step": 104310 + }, + { + "epoch": 30.352051207448355, + "grad_norm": 0.15211069583892822, + "learning_rate": 1.3788909406910927e-06, + "loss": 0.0113, + "step": 104320 + }, + { + "epoch": 30.3549607215595, + "grad_norm": 0.12965542078018188, + "learning_rate": 1.3771602341462664e-06, + "loss": 0.0115, + "step": 104330 + }, + { + "epoch": 30.357870235670642, + "grad_norm": 0.11057061702013016, + "learning_rate": 1.3754305621662138e-06, + "loss": 0.0109, + "step": 104340 + }, + { + "epoch": 30.360779749781788, + "grad_norm": 0.12737445533275604, + "learning_rate": 1.3737019248823e-06, + "loss": 0.0108, + "step": 104350 + }, + { + "epoch": 30.36368926389293, + "grad_norm": 0.07845453917980194, + "learning_rate": 1.3719743224257986e-06, + "loss": 0.0097, + "step": 104360 + }, + { + "epoch": 30.36659877800407, + "grad_norm": 0.09377983212471008, + "learning_rate": 1.3702477549279092e-06, + "loss": 0.0097, + "step": 104370 + }, + { + "epoch": 30.369508292115217, + "grad_norm": 0.13006684184074402, + "learning_rate": 1.3685222225197536e-06, + "loss": 0.0112, + "step": 104380 + }, + { + "epoch": 30.37241780622636, + "grad_norm": 0.127424955368042, + "learning_rate": 1.366797725332377e-06, + "loss": 0.0109, + "step": 104390 + }, + { + "epoch": 30.375327320337504, + "grad_norm": 0.07106469571590424, + "learning_rate": 1.3650742634967433e-06, + "loss": 0.0104, + "step": 104400 + }, + { + "epoch": 30.378236834448646, + "grad_norm": 0.15895120799541473, + "learning_rate": 1.3633518371437326e-06, + "loss": 0.0125, + "step": 104410 + }, + { + "epoch": 30.381146348559792, + "grad_norm": 0.05815383046865463, + "learning_rate": 1.3616304464041568e-06, + "loss": 0.0104, + "step": 104420 + }, + { + "epoch": 30.384055862670934, + "grad_norm": 0.12298205494880676, + "learning_rate": 1.3599100914087465e-06, + "loss": 0.0094, + "step": 104430 + }, + { + "epoch": 30.386965376782076, + "grad_norm": 0.19083473086357117, + "learning_rate": 1.3581907722881436e-06, + "loss": 0.0108, + "step": 104440 + }, + { + "epoch": 30.38987489089322, + "grad_norm": 0.1595262885093689, + "learning_rate": 1.356472489172924e-06, + "loss": 0.0105, + "step": 104450 + }, + { + "epoch": 30.392784405004363, + "grad_norm": 0.15094433724880219, + "learning_rate": 1.3547552421935832e-06, + "loss": 0.0113, + "step": 104460 + }, + { + "epoch": 30.39569391911551, + "grad_norm": 0.213905930519104, + "learning_rate": 1.3530390314805303e-06, + "loss": 0.0104, + "step": 104470 + }, + { + "epoch": 30.39860343322665, + "grad_norm": 0.08591470122337341, + "learning_rate": 1.3513238571640973e-06, + "loss": 0.0094, + "step": 104480 + }, + { + "epoch": 30.401512947337796, + "grad_norm": 0.09995967149734497, + "learning_rate": 1.3496097193745437e-06, + "loss": 0.0097, + "step": 104490 + }, + { + "epoch": 30.404422461448938, + "grad_norm": 0.11659444868564606, + "learning_rate": 1.3478966182420522e-06, + "loss": 0.0115, + "step": 104500 + }, + { + "epoch": 30.40733197556008, + "grad_norm": 0.10706035047769547, + "learning_rate": 1.3461845538967115e-06, + "loss": 0.0103, + "step": 104510 + }, + { + "epoch": 30.410241489671225, + "grad_norm": 0.13555824756622314, + "learning_rate": 1.3444735264685497e-06, + "loss": 0.0104, + "step": 104520 + }, + { + "epoch": 30.413151003782367, + "grad_norm": 0.11128878593444824, + "learning_rate": 1.342763536087503e-06, + "loss": 0.01, + "step": 104530 + }, + { + "epoch": 30.416060517893513, + "grad_norm": 0.13133321702480316, + "learning_rate": 1.3410545828834352e-06, + "loss": 0.0095, + "step": 104540 + }, + { + "epoch": 30.418970032004655, + "grad_norm": 0.10979580879211426, + "learning_rate": 1.339346666986135e-06, + "loss": 0.0107, + "step": 104550 + }, + { + "epoch": 30.4218795461158, + "grad_norm": 0.13786691427230835, + "learning_rate": 1.3376397885252977e-06, + "loss": 0.0119, + "step": 104560 + }, + { + "epoch": 30.424789060226942, + "grad_norm": 0.1364724040031433, + "learning_rate": 1.3359339476305589e-06, + "loss": 0.0122, + "step": 104570 + }, + { + "epoch": 30.427698574338084, + "grad_norm": 0.14525219798088074, + "learning_rate": 1.3342291444314592e-06, + "loss": 0.0098, + "step": 104580 + }, + { + "epoch": 30.43060808844923, + "grad_norm": 0.09344445914030075, + "learning_rate": 1.332525379057472e-06, + "loss": 0.0098, + "step": 104590 + }, + { + "epoch": 30.43351760256037, + "grad_norm": 0.133214071393013, + "learning_rate": 1.3308226516379824e-06, + "loss": 0.0111, + "step": 104600 + }, + { + "epoch": 30.436427116671517, + "grad_norm": 0.1435423046350479, + "learning_rate": 1.3291209623023032e-06, + "loss": 0.0098, + "step": 104610 + }, + { + "epoch": 30.43933663078266, + "grad_norm": 0.10436933487653732, + "learning_rate": 1.3274203111796695e-06, + "loss": 0.0078, + "step": 104620 + }, + { + "epoch": 30.442246144893804, + "grad_norm": 0.11242033541202545, + "learning_rate": 1.325720698399233e-06, + "loss": 0.01, + "step": 104630 + }, + { + "epoch": 30.445155659004946, + "grad_norm": 0.11749500036239624, + "learning_rate": 1.3240221240900607e-06, + "loss": 0.0093, + "step": 104640 + }, + { + "epoch": 30.44806517311609, + "grad_norm": 0.11556417495012283, + "learning_rate": 1.3223245883811592e-06, + "loss": 0.0132, + "step": 104650 + }, + { + "epoch": 30.450974687227234, + "grad_norm": 0.08100299537181854, + "learning_rate": 1.3206280914014418e-06, + "loss": 0.0093, + "step": 104660 + }, + { + "epoch": 30.453884201338376, + "grad_norm": 0.10251624137163162, + "learning_rate": 1.3189326332797425e-06, + "loss": 0.0106, + "step": 104670 + }, + { + "epoch": 30.45679371544952, + "grad_norm": 0.10258842259645462, + "learning_rate": 1.3172382141448214e-06, + "loss": 0.0107, + "step": 104680 + }, + { + "epoch": 30.459703229560663, + "grad_norm": 0.08390691876411438, + "learning_rate": 1.3155448341253623e-06, + "loss": 0.0111, + "step": 104690 + }, + { + "epoch": 30.462612743671805, + "grad_norm": 0.07769119739532471, + "learning_rate": 1.3138524933499641e-06, + "loss": 0.0114, + "step": 104700 + }, + { + "epoch": 30.46552225778295, + "grad_norm": 0.11187168955802917, + "learning_rate": 1.3121611919471455e-06, + "loss": 0.0119, + "step": 104710 + }, + { + "epoch": 30.468431771894092, + "grad_norm": 0.1140156090259552, + "learning_rate": 1.310470930045352e-06, + "loss": 0.0126, + "step": 104720 + }, + { + "epoch": 30.471341286005238, + "grad_norm": 0.08819785714149475, + "learning_rate": 1.3087817077729513e-06, + "loss": 0.0128, + "step": 104730 + }, + { + "epoch": 30.47425080011638, + "grad_norm": 0.10603176802396774, + "learning_rate": 1.307093525258224e-06, + "loss": 0.0122, + "step": 104740 + }, + { + "epoch": 30.477160314227525, + "grad_norm": 0.08151881396770477, + "learning_rate": 1.3054063826293771e-06, + "loss": 0.0089, + "step": 104750 + }, + { + "epoch": 30.480069828338667, + "grad_norm": 0.14047616720199585, + "learning_rate": 1.303720280014542e-06, + "loss": 0.0122, + "step": 104760 + }, + { + "epoch": 30.48297934244981, + "grad_norm": 0.10781526565551758, + "learning_rate": 1.3020352175417626e-06, + "loss": 0.0101, + "step": 104770 + }, + { + "epoch": 30.485888856560955, + "grad_norm": 0.1114276573061943, + "learning_rate": 1.3003511953390084e-06, + "loss": 0.0085, + "step": 104780 + }, + { + "epoch": 30.488798370672097, + "grad_norm": 0.12103437632322311, + "learning_rate": 1.2986682135341722e-06, + "loss": 0.0107, + "step": 104790 + }, + { + "epoch": 30.491707884783242, + "grad_norm": 0.09844201803207397, + "learning_rate": 1.2969862722550647e-06, + "loss": 0.0101, + "step": 104800 + }, + { + "epoch": 30.494617398894384, + "grad_norm": 0.1580093502998352, + "learning_rate": 1.2953053716294173e-06, + "loss": 0.0103, + "step": 104810 + }, + { + "epoch": 30.49752691300553, + "grad_norm": 0.12066905200481415, + "learning_rate": 1.2936255117848843e-06, + "loss": 0.0101, + "step": 104820 + }, + { + "epoch": 30.50043642711667, + "grad_norm": 0.10575578361749649, + "learning_rate": 1.2919466928490392e-06, + "loss": 0.0101, + "step": 104830 + }, + { + "epoch": 30.503345941227813, + "grad_norm": 0.08909182250499725, + "learning_rate": 1.2902689149493779e-06, + "loss": 0.0114, + "step": 104840 + }, + { + "epoch": 30.50625545533896, + "grad_norm": 0.11060108989477158, + "learning_rate": 1.2885921782133153e-06, + "loss": 0.0112, + "step": 104850 + }, + { + "epoch": 30.5091649694501, + "grad_norm": 0.1166629046201706, + "learning_rate": 1.286916482768188e-06, + "loss": 0.0108, + "step": 104860 + }, + { + "epoch": 30.512074483561246, + "grad_norm": 0.07995465397834778, + "learning_rate": 1.2852418287412588e-06, + "loss": 0.0097, + "step": 104870 + }, + { + "epoch": 30.51498399767239, + "grad_norm": 0.10147503018379211, + "learning_rate": 1.2835682162596997e-06, + "loss": 0.013, + "step": 104880 + }, + { + "epoch": 30.517893511783534, + "grad_norm": 0.13056348264217377, + "learning_rate": 1.281895645450617e-06, + "loss": 0.0097, + "step": 104890 + }, + { + "epoch": 30.520803025894676, + "grad_norm": 0.1107187420129776, + "learning_rate": 1.2802241164410262e-06, + "loss": 0.0095, + "step": 104900 + }, + { + "epoch": 30.523712540005818, + "grad_norm": 0.15526622533798218, + "learning_rate": 1.2785536293578703e-06, + "loss": 0.0105, + "step": 104910 + }, + { + "epoch": 30.526622054116963, + "grad_norm": 0.07806511968374252, + "learning_rate": 1.2768841843280165e-06, + "loss": 0.0104, + "step": 104920 + }, + { + "epoch": 30.529531568228105, + "grad_norm": 0.2299979031085968, + "learning_rate": 1.2752157814782433e-06, + "loss": 0.011, + "step": 104930 + }, + { + "epoch": 30.53244108233925, + "grad_norm": 0.09685398638248444, + "learning_rate": 1.2735484209352528e-06, + "loss": 0.0115, + "step": 104940 + }, + { + "epoch": 30.535350596450392, + "grad_norm": 0.10169509798288345, + "learning_rate": 1.271882102825675e-06, + "loss": 0.0095, + "step": 104950 + }, + { + "epoch": 30.538260110561538, + "grad_norm": 0.08994588255882263, + "learning_rate": 1.2702168272760538e-06, + "loss": 0.0108, + "step": 104960 + }, + { + "epoch": 30.54116962467268, + "grad_norm": 0.38209179043769836, + "learning_rate": 1.268552594412855e-06, + "loss": 0.0103, + "step": 104970 + }, + { + "epoch": 30.54407913878382, + "grad_norm": 0.152121901512146, + "learning_rate": 1.2668894043624657e-06, + "loss": 0.0117, + "step": 104980 + }, + { + "epoch": 30.546988652894967, + "grad_norm": 0.09875912219285965, + "learning_rate": 1.2652272572511998e-06, + "loss": 0.0131, + "step": 104990 + }, + { + "epoch": 30.54989816700611, + "grad_norm": 0.1176685020327568, + "learning_rate": 1.2635661532052812e-06, + "loss": 0.0115, + "step": 105000 + }, + { + "epoch": 30.552807681117255, + "grad_norm": 0.1314648687839508, + "learning_rate": 1.2619060923508574e-06, + "loss": 0.0098, + "step": 105010 + }, + { + "epoch": 30.555717195228397, + "grad_norm": 0.22043214738368988, + "learning_rate": 1.2602470748140028e-06, + "loss": 0.0097, + "step": 105020 + }, + { + "epoch": 30.558626709339542, + "grad_norm": 0.13740983605384827, + "learning_rate": 1.2585891007207113e-06, + "loss": 0.0098, + "step": 105030 + }, + { + "epoch": 30.561536223450684, + "grad_norm": 0.08657616376876831, + "learning_rate": 1.2569321701968922e-06, + "loss": 0.0098, + "step": 105040 + }, + { + "epoch": 30.564445737561826, + "grad_norm": 0.11109429597854614, + "learning_rate": 1.2552762833683735e-06, + "loss": 0.0106, + "step": 105050 + }, + { + "epoch": 30.56735525167297, + "grad_norm": 0.07598654180765152, + "learning_rate": 1.2536214403609193e-06, + "loss": 0.0088, + "step": 105060 + }, + { + "epoch": 30.570264765784113, + "grad_norm": 0.0785161480307579, + "learning_rate": 1.2519676413001974e-06, + "loss": 0.0103, + "step": 105070 + }, + { + "epoch": 30.57317427989526, + "grad_norm": 0.1078147143125534, + "learning_rate": 1.2503148863118008e-06, + "loss": 0.01, + "step": 105080 + }, + { + "epoch": 30.5760837940064, + "grad_norm": 0.08934729546308517, + "learning_rate": 1.2486631755212507e-06, + "loss": 0.0103, + "step": 105090 + }, + { + "epoch": 30.578993308117543, + "grad_norm": 0.09432261437177658, + "learning_rate": 1.2470125090539819e-06, + "loss": 0.01, + "step": 105100 + }, + { + "epoch": 30.581902822228688, + "grad_norm": 0.08351969718933105, + "learning_rate": 1.2453628870353506e-06, + "loss": 0.0095, + "step": 105110 + }, + { + "epoch": 30.58481233633983, + "grad_norm": 0.129583477973938, + "learning_rate": 1.2437143095906367e-06, + "loss": 0.0138, + "step": 105120 + }, + { + "epoch": 30.587721850450976, + "grad_norm": 0.08456286042928696, + "learning_rate": 1.242066776845035e-06, + "loss": 0.0095, + "step": 105130 + }, + { + "epoch": 30.590631364562118, + "grad_norm": 0.10005814582109451, + "learning_rate": 1.2404202889236705e-06, + "loss": 0.0113, + "step": 105140 + }, + { + "epoch": 30.593540878673263, + "grad_norm": 0.11066337674856186, + "learning_rate": 1.2387748459515763e-06, + "loss": 0.0089, + "step": 105150 + }, + { + "epoch": 30.596450392784405, + "grad_norm": 0.2754073143005371, + "learning_rate": 1.237130448053716e-06, + "loss": 0.0097, + "step": 105160 + }, + { + "epoch": 30.599359906895547, + "grad_norm": 0.09211724996566772, + "learning_rate": 1.2354870953549745e-06, + "loss": 0.0096, + "step": 105170 + }, + { + "epoch": 30.602269421006692, + "grad_norm": 0.1400904506444931, + "learning_rate": 1.2338447879801472e-06, + "loss": 0.0116, + "step": 105180 + }, + { + "epoch": 30.605178935117834, + "grad_norm": 0.12306106090545654, + "learning_rate": 1.2322035260539627e-06, + "loss": 0.012, + "step": 105190 + }, + { + "epoch": 30.60808844922898, + "grad_norm": 0.06914076954126358, + "learning_rate": 1.2305633097010577e-06, + "loss": 0.0106, + "step": 105200 + }, + { + "epoch": 30.61099796334012, + "grad_norm": 0.12039081007242203, + "learning_rate": 1.2289241390459994e-06, + "loss": 0.0112, + "step": 105210 + }, + { + "epoch": 30.613907477451267, + "grad_norm": 0.0933828130364418, + "learning_rate": 1.2272860142132731e-06, + "loss": 0.0102, + "step": 105220 + }, + { + "epoch": 30.61681699156241, + "grad_norm": 0.1234767884016037, + "learning_rate": 1.2256489353272843e-06, + "loss": 0.0099, + "step": 105230 + }, + { + "epoch": 30.61972650567355, + "grad_norm": 0.10486501455307007, + "learning_rate": 1.224012902512352e-06, + "loss": 0.0108, + "step": 105240 + }, + { + "epoch": 30.622636019784697, + "grad_norm": 0.11062643676996231, + "learning_rate": 1.222377915892725e-06, + "loss": 0.0088, + "step": 105250 + }, + { + "epoch": 30.62554553389584, + "grad_norm": 0.09878414124250412, + "learning_rate": 1.2207439755925743e-06, + "loss": 0.0104, + "step": 105260 + }, + { + "epoch": 30.628455048006984, + "grad_norm": 0.08611579984426498, + "learning_rate": 1.2191110817359818e-06, + "loss": 0.0112, + "step": 105270 + }, + { + "epoch": 30.631364562118126, + "grad_norm": 0.12052283436059952, + "learning_rate": 1.2174792344469554e-06, + "loss": 0.0103, + "step": 105280 + }, + { + "epoch": 30.63427407622927, + "grad_norm": 0.08477263897657394, + "learning_rate": 1.215848433849429e-06, + "loss": 0.0087, + "step": 105290 + }, + { + "epoch": 30.637183590340413, + "grad_norm": 0.10584204643964767, + "learning_rate": 1.214218680067244e-06, + "loss": 0.0112, + "step": 105300 + }, + { + "epoch": 30.640093104451555, + "grad_norm": 0.17931678891181946, + "learning_rate": 1.2125899732241709e-06, + "loss": 0.0115, + "step": 105310 + }, + { + "epoch": 30.6430026185627, + "grad_norm": 0.09551700949668884, + "learning_rate": 1.2109623134439007e-06, + "loss": 0.0103, + "step": 105320 + }, + { + "epoch": 30.645912132673843, + "grad_norm": 0.1928028166294098, + "learning_rate": 1.209335700850045e-06, + "loss": 0.0112, + "step": 105330 + }, + { + "epoch": 30.648821646784988, + "grad_norm": 0.1966690868139267, + "learning_rate": 1.207710135566133e-06, + "loss": 0.0117, + "step": 105340 + }, + { + "epoch": 30.65173116089613, + "grad_norm": 0.09361927211284637, + "learning_rate": 1.2060856177156092e-06, + "loss": 0.0092, + "step": 105350 + }, + { + "epoch": 30.654640675007272, + "grad_norm": 0.09442639350891113, + "learning_rate": 1.2044621474218553e-06, + "loss": 0.0115, + "step": 105360 + }, + { + "epoch": 30.657550189118417, + "grad_norm": 0.1732500046491623, + "learning_rate": 1.202839724808159e-06, + "loss": 0.0088, + "step": 105370 + }, + { + "epoch": 30.66045970322956, + "grad_norm": 0.11532370001077652, + "learning_rate": 1.2012183499977303e-06, + "loss": 0.0089, + "step": 105380 + }, + { + "epoch": 30.663369217340705, + "grad_norm": 0.0879669338464737, + "learning_rate": 1.1995980231137038e-06, + "loss": 0.0116, + "step": 105390 + }, + { + "epoch": 30.666278731451847, + "grad_norm": 0.0995737612247467, + "learning_rate": 1.1979787442791363e-06, + "loss": 0.0121, + "step": 105400 + }, + { + "epoch": 30.669188245562992, + "grad_norm": 0.10677237808704376, + "learning_rate": 1.1963605136169959e-06, + "loss": 0.0109, + "step": 105410 + }, + { + "epoch": 30.672097759674134, + "grad_norm": 0.17473742365837097, + "learning_rate": 1.194743331250176e-06, + "loss": 0.0104, + "step": 105420 + }, + { + "epoch": 30.675007273785276, + "grad_norm": 0.1314888447523117, + "learning_rate": 1.1931271973014952e-06, + "loss": 0.0121, + "step": 105430 + }, + { + "epoch": 30.67791678789642, + "grad_norm": 0.16882877051830292, + "learning_rate": 1.1915121118936866e-06, + "loss": 0.0084, + "step": 105440 + }, + { + "epoch": 30.680826302007564, + "grad_norm": 0.06116114556789398, + "learning_rate": 1.1898980751494043e-06, + "loss": 0.0103, + "step": 105450 + }, + { + "epoch": 30.68373581611871, + "grad_norm": 0.11242896318435669, + "learning_rate": 1.1882850871912231e-06, + "loss": 0.0112, + "step": 105460 + }, + { + "epoch": 30.68664533022985, + "grad_norm": 0.10195382684469223, + "learning_rate": 1.1866731481416438e-06, + "loss": 0.0101, + "step": 105470 + }, + { + "epoch": 30.689554844340996, + "grad_norm": 0.10099457949399948, + "learning_rate": 1.1850622581230747e-06, + "loss": 0.009, + "step": 105480 + }, + { + "epoch": 30.69246435845214, + "grad_norm": 0.12538045644760132, + "learning_rate": 1.1834524172578598e-06, + "loss": 0.0116, + "step": 105490 + }, + { + "epoch": 30.69537387256328, + "grad_norm": 0.1501418501138687, + "learning_rate": 1.1818436256682514e-06, + "loss": 0.01, + "step": 105500 + }, + { + "epoch": 30.698283386674426, + "grad_norm": 0.12243220955133438, + "learning_rate": 1.1802358834764283e-06, + "loss": 0.0117, + "step": 105510 + }, + { + "epoch": 30.701192900785568, + "grad_norm": 0.09787268191576004, + "learning_rate": 1.178629190804486e-06, + "loss": 0.012, + "step": 105520 + }, + { + "epoch": 30.704102414896713, + "grad_norm": 0.16230736672878265, + "learning_rate": 1.1770235477744457e-06, + "loss": 0.0126, + "step": 105530 + }, + { + "epoch": 30.707011929007855, + "grad_norm": 0.11497882753610611, + "learning_rate": 1.1754189545082394e-06, + "loss": 0.0098, + "step": 105540 + }, + { + "epoch": 30.709921443119, + "grad_norm": 0.11968881636857986, + "learning_rate": 1.1738154111277317e-06, + "loss": 0.0086, + "step": 105550 + }, + { + "epoch": 30.712830957230143, + "grad_norm": 0.1798691749572754, + "learning_rate": 1.1722129177546997e-06, + "loss": 0.0099, + "step": 105560 + }, + { + "epoch": 30.715740471341284, + "grad_norm": 0.1273152083158493, + "learning_rate": 1.1706114745108381e-06, + "loss": 0.0113, + "step": 105570 + }, + { + "epoch": 30.71864998545243, + "grad_norm": 0.11380680650472641, + "learning_rate": 1.1690110815177697e-06, + "loss": 0.0105, + "step": 105580 + }, + { + "epoch": 30.721559499563572, + "grad_norm": 0.07267022132873535, + "learning_rate": 1.1674117388970367e-06, + "loss": 0.0103, + "step": 105590 + }, + { + "epoch": 30.724469013674717, + "grad_norm": 0.11756256222724915, + "learning_rate": 1.1658134467700927e-06, + "loss": 0.0111, + "step": 105600 + }, + { + "epoch": 30.72737852778586, + "grad_norm": 0.0957062691450119, + "learning_rate": 1.1642162052583182e-06, + "loss": 0.0097, + "step": 105610 + }, + { + "epoch": 30.730288041897005, + "grad_norm": 0.1110597550868988, + "learning_rate": 1.162620014483015e-06, + "loss": 0.0108, + "step": 105620 + }, + { + "epoch": 30.733197556008147, + "grad_norm": 0.06420084089040756, + "learning_rate": 1.1610248745654061e-06, + "loss": 0.011, + "step": 105630 + }, + { + "epoch": 30.73610707011929, + "grad_norm": 0.08937204629182816, + "learning_rate": 1.1594307856266278e-06, + "loss": 0.0096, + "step": 105640 + }, + { + "epoch": 30.739016584230434, + "grad_norm": 0.08628872781991959, + "learning_rate": 1.1578377477877366e-06, + "loss": 0.0109, + "step": 105650 + }, + { + "epoch": 30.741926098341576, + "grad_norm": 0.1813594251871109, + "learning_rate": 1.1562457611697241e-06, + "loss": 0.0114, + "step": 105660 + }, + { + "epoch": 30.74483561245272, + "grad_norm": 0.1289631873369217, + "learning_rate": 1.1546548258934837e-06, + "loss": 0.0107, + "step": 105670 + }, + { + "epoch": 30.747745126563863, + "grad_norm": 0.12353561818599701, + "learning_rate": 1.1530649420798368e-06, + "loss": 0.0092, + "step": 105680 + }, + { + "epoch": 30.75065464067501, + "grad_norm": 0.10520146787166595, + "learning_rate": 1.1514761098495236e-06, + "loss": 0.0113, + "step": 105690 + }, + { + "epoch": 30.75356415478615, + "grad_norm": 0.0957864448428154, + "learning_rate": 1.1498883293232126e-06, + "loss": 0.0117, + "step": 105700 + }, + { + "epoch": 30.756473668897293, + "grad_norm": 0.10428518056869507, + "learning_rate": 1.1483016006214793e-06, + "loss": 0.009, + "step": 105710 + }, + { + "epoch": 30.75938318300844, + "grad_norm": 0.1630682796239853, + "learning_rate": 1.14671592386482e-06, + "loss": 0.0103, + "step": 105720 + }, + { + "epoch": 30.76229269711958, + "grad_norm": 0.05504153296351433, + "learning_rate": 1.145131299173669e-06, + "loss": 0.0105, + "step": 105730 + }, + { + "epoch": 30.765202211230726, + "grad_norm": 0.15049174427986145, + "learning_rate": 1.1435477266683614e-06, + "loss": 0.0101, + "step": 105740 + }, + { + "epoch": 30.768111725341868, + "grad_norm": 0.1247660219669342, + "learning_rate": 1.1419652064691561e-06, + "loss": 0.01, + "step": 105750 + }, + { + "epoch": 30.771021239453013, + "grad_norm": 0.08731770515441895, + "learning_rate": 1.1403837386962384e-06, + "loss": 0.0093, + "step": 105760 + }, + { + "epoch": 30.773930753564155, + "grad_norm": 0.07357361167669296, + "learning_rate": 1.1388033234697127e-06, + "loss": 0.009, + "step": 105770 + }, + { + "epoch": 30.776840267675297, + "grad_norm": 0.11443887650966644, + "learning_rate": 1.1372239609095992e-06, + "loss": 0.013, + "step": 105780 + }, + { + "epoch": 30.779749781786442, + "grad_norm": 0.09305381029844284, + "learning_rate": 1.135645651135837e-06, + "loss": 0.0096, + "step": 105790 + }, + { + "epoch": 30.782659295897584, + "grad_norm": 0.1310179978609085, + "learning_rate": 1.1340683942682905e-06, + "loss": 0.0104, + "step": 105800 + }, + { + "epoch": 30.78556881000873, + "grad_norm": 0.09404555708169937, + "learning_rate": 1.1324921904267454e-06, + "loss": 0.0097, + "step": 105810 + }, + { + "epoch": 30.788478324119872, + "grad_norm": 0.1702103316783905, + "learning_rate": 1.1309170397308977e-06, + "loss": 0.011, + "step": 105820 + }, + { + "epoch": 30.791387838231014, + "grad_norm": 0.07570072263479233, + "learning_rate": 1.1293429423003732e-06, + "loss": 0.0095, + "step": 105830 + }, + { + "epoch": 30.79429735234216, + "grad_norm": 0.09993129968643188, + "learning_rate": 1.1277698982547162e-06, + "loss": 0.0079, + "step": 105840 + }, + { + "epoch": 30.7972068664533, + "grad_norm": 0.18619219958782196, + "learning_rate": 1.126197907713385e-06, + "loss": 0.0107, + "step": 105850 + }, + { + "epoch": 30.800116380564447, + "grad_norm": 0.0764448493719101, + "learning_rate": 1.1246269707957652e-06, + "loss": 0.0101, + "step": 105860 + }, + { + "epoch": 30.80302589467559, + "grad_norm": 0.1756758689880371, + "learning_rate": 1.1230570876211549e-06, + "loss": 0.0114, + "step": 105870 + }, + { + "epoch": 30.805935408786734, + "grad_norm": 0.1645403951406479, + "learning_rate": 1.1214882583087805e-06, + "loss": 0.011, + "step": 105880 + }, + { + "epoch": 30.808844922897876, + "grad_norm": 0.1125633716583252, + "learning_rate": 1.1199204829777814e-06, + "loss": 0.0108, + "step": 105890 + }, + { + "epoch": 30.811754437009018, + "grad_norm": 0.11223208159208298, + "learning_rate": 1.1183537617472246e-06, + "loss": 0.0106, + "step": 105900 + }, + { + "epoch": 30.814663951120163, + "grad_norm": 0.0834275558590889, + "learning_rate": 1.1167880947360842e-06, + "loss": 0.0107, + "step": 105910 + }, + { + "epoch": 30.817573465231305, + "grad_norm": 0.08721768110990524, + "learning_rate": 1.1152234820632672e-06, + "loss": 0.0116, + "step": 105920 + }, + { + "epoch": 30.82048297934245, + "grad_norm": 0.08290717005729675, + "learning_rate": 1.1136599238475986e-06, + "loss": 0.0105, + "step": 105930 + }, + { + "epoch": 30.823392493453593, + "grad_norm": 0.08422444760799408, + "learning_rate": 1.1120974202078133e-06, + "loss": 0.0098, + "step": 105940 + }, + { + "epoch": 30.82630200756474, + "grad_norm": 0.08187083899974823, + "learning_rate": 1.1105359712625778e-06, + "loss": 0.0098, + "step": 105950 + }, + { + "epoch": 30.82921152167588, + "grad_norm": 0.0955839604139328, + "learning_rate": 1.1089755771304743e-06, + "loss": 0.0098, + "step": 105960 + }, + { + "epoch": 30.832121035787022, + "grad_norm": 0.10452260076999664, + "learning_rate": 1.1074162379300046e-06, + "loss": 0.0106, + "step": 105970 + }, + { + "epoch": 30.835030549898168, + "grad_norm": 0.14107225835323334, + "learning_rate": 1.1058579537795838e-06, + "loss": 0.0112, + "step": 105980 + }, + { + "epoch": 30.83794006400931, + "grad_norm": 0.10060081630945206, + "learning_rate": 1.1043007247975605e-06, + "loss": 0.009, + "step": 105990 + }, + { + "epoch": 30.840849578120455, + "grad_norm": 0.11708849668502808, + "learning_rate": 1.1027445511021939e-06, + "loss": 0.01, + "step": 106000 + }, + { + "epoch": 30.843759092231597, + "grad_norm": 0.14106833934783936, + "learning_rate": 1.1011894328116656e-06, + "loss": 0.0116, + "step": 106010 + }, + { + "epoch": 30.846668606342742, + "grad_norm": 0.1325356811285019, + "learning_rate": 1.0996353700440713e-06, + "loss": 0.0118, + "step": 106020 + }, + { + "epoch": 30.849578120453884, + "grad_norm": 0.142317995429039, + "learning_rate": 1.0980823629174403e-06, + "loss": 0.0111, + "step": 106030 + }, + { + "epoch": 30.852487634565026, + "grad_norm": 0.10578326880931854, + "learning_rate": 1.0965304115497094e-06, + "loss": 0.0111, + "step": 106040 + }, + { + "epoch": 30.855397148676172, + "grad_norm": 0.1172722652554512, + "learning_rate": 1.0949795160587361e-06, + "loss": 0.0113, + "step": 106050 + }, + { + "epoch": 30.858306662787314, + "grad_norm": 0.13601167500019073, + "learning_rate": 1.093429676562303e-06, + "loss": 0.0103, + "step": 106060 + }, + { + "epoch": 30.86121617689846, + "grad_norm": 0.14073127508163452, + "learning_rate": 1.0918808931781144e-06, + "loss": 0.0124, + "step": 106070 + }, + { + "epoch": 30.8641256910096, + "grad_norm": 0.07986371219158173, + "learning_rate": 1.090333166023786e-06, + "loss": 0.011, + "step": 106080 + }, + { + "epoch": 30.867035205120743, + "grad_norm": 0.0914507582783699, + "learning_rate": 1.0887864952168554e-06, + "loss": 0.0133, + "step": 106090 + }, + { + "epoch": 30.86994471923189, + "grad_norm": 0.06820007413625717, + "learning_rate": 1.0872408808747837e-06, + "loss": 0.0102, + "step": 106100 + }, + { + "epoch": 30.87285423334303, + "grad_norm": 0.06841875612735748, + "learning_rate": 1.0856963231149535e-06, + "loss": 0.0101, + "step": 106110 + }, + { + "epoch": 30.875763747454176, + "grad_norm": 0.14162376523017883, + "learning_rate": 1.0841528220546598e-06, + "loss": 0.0124, + "step": 106120 + }, + { + "epoch": 30.878673261565318, + "grad_norm": 0.0794142335653305, + "learning_rate": 1.0826103778111218e-06, + "loss": 0.0104, + "step": 106130 + }, + { + "epoch": 30.881582775676463, + "grad_norm": 0.07862628251314163, + "learning_rate": 1.0810689905014808e-06, + "loss": 0.0102, + "step": 106140 + }, + { + "epoch": 30.884492289787605, + "grad_norm": 0.11442065984010696, + "learning_rate": 1.0795286602427933e-06, + "loss": 0.0108, + "step": 106150 + }, + { + "epoch": 30.887401803898747, + "grad_norm": 0.10009769350290298, + "learning_rate": 1.0779893871520342e-06, + "loss": 0.0111, + "step": 106160 + }, + { + "epoch": 30.890311318009893, + "grad_norm": 0.1168542355298996, + "learning_rate": 1.0764511713461044e-06, + "loss": 0.0098, + "step": 106170 + }, + { + "epoch": 30.893220832121035, + "grad_norm": 0.18982963263988495, + "learning_rate": 1.0749140129418228e-06, + "loss": 0.0118, + "step": 106180 + }, + { + "epoch": 30.89613034623218, + "grad_norm": 0.09686245024204254, + "learning_rate": 1.0733779120559223e-06, + "loss": 0.009, + "step": 106190 + }, + { + "epoch": 30.899039860343322, + "grad_norm": 0.17040592432022095, + "learning_rate": 1.0718428688050647e-06, + "loss": 0.0118, + "step": 106200 + }, + { + "epoch": 30.901949374454468, + "grad_norm": 0.08529935777187347, + "learning_rate": 1.07030888330582e-06, + "loss": 0.01, + "step": 106210 + }, + { + "epoch": 30.90485888856561, + "grad_norm": 0.08268829435110092, + "learning_rate": 1.0687759556746884e-06, + "loss": 0.0089, + "step": 106220 + }, + { + "epoch": 30.90776840267675, + "grad_norm": 0.08449345082044601, + "learning_rate": 1.067244086028087e-06, + "loss": 0.0098, + "step": 106230 + }, + { + "epoch": 30.910677916787897, + "grad_norm": 0.09079550206661224, + "learning_rate": 1.0657132744823456e-06, + "loss": 0.0083, + "step": 106240 + }, + { + "epoch": 30.91358743089904, + "grad_norm": 0.129109188914299, + "learning_rate": 1.0641835211537265e-06, + "loss": 0.0081, + "step": 106250 + }, + { + "epoch": 30.916496945010184, + "grad_norm": 0.10426995903253555, + "learning_rate": 1.0626548261583986e-06, + "loss": 0.0102, + "step": 106260 + }, + { + "epoch": 30.919406459121326, + "grad_norm": 0.08247467875480652, + "learning_rate": 1.0611271896124585e-06, + "loss": 0.009, + "step": 106270 + }, + { + "epoch": 30.92231597323247, + "grad_norm": 0.08955986052751541, + "learning_rate": 1.059600611631919e-06, + "loss": 0.0102, + "step": 106280 + }, + { + "epoch": 30.925225487343614, + "grad_norm": 0.20500734448432922, + "learning_rate": 1.0580750923327136e-06, + "loss": 0.0123, + "step": 106290 + }, + { + "epoch": 30.928135001454756, + "grad_norm": 0.10324487835168839, + "learning_rate": 1.0565506318307e-06, + "loss": 0.0104, + "step": 106300 + }, + { + "epoch": 30.9310445155659, + "grad_norm": 0.20457687973976135, + "learning_rate": 1.0550272302416453e-06, + "loss": 0.0091, + "step": 106310 + }, + { + "epoch": 30.933954029677043, + "grad_norm": 0.15484948456287384, + "learning_rate": 1.0535048876812402e-06, + "loss": 0.0096, + "step": 106320 + }, + { + "epoch": 30.93686354378819, + "grad_norm": 0.0992451161146164, + "learning_rate": 1.0519836042651042e-06, + "loss": 0.0117, + "step": 106330 + }, + { + "epoch": 30.93977305789933, + "grad_norm": 0.11159677803516388, + "learning_rate": 1.0504633801087648e-06, + "loss": 0.0099, + "step": 106340 + }, + { + "epoch": 30.942682572010476, + "grad_norm": 0.09494244307279587, + "learning_rate": 1.0489442153276696e-06, + "loss": 0.0088, + "step": 106350 + }, + { + "epoch": 30.945592086121618, + "grad_norm": 0.11522167921066284, + "learning_rate": 1.047426110037193e-06, + "loss": 0.0098, + "step": 106360 + }, + { + "epoch": 30.94850160023276, + "grad_norm": 0.1437743902206421, + "learning_rate": 1.0459090643526275e-06, + "loss": 0.0104, + "step": 106370 + }, + { + "epoch": 30.951411114343905, + "grad_norm": 0.11529731750488281, + "learning_rate": 1.0443930783891782e-06, + "loss": 0.0142, + "step": 106380 + }, + { + "epoch": 30.954320628455047, + "grad_norm": 0.07113231718540192, + "learning_rate": 1.0428781522619741e-06, + "loss": 0.0093, + "step": 106390 + }, + { + "epoch": 30.957230142566193, + "grad_norm": 0.1055554449558258, + "learning_rate": 1.041364286086065e-06, + "loss": 0.0088, + "step": 106400 + }, + { + "epoch": 30.960139656677335, + "grad_norm": 0.08658762276172638, + "learning_rate": 1.0398514799764226e-06, + "loss": 0.0094, + "step": 106410 + }, + { + "epoch": 30.96304917078848, + "grad_norm": 0.11672456562519073, + "learning_rate": 1.0383397340479279e-06, + "loss": 0.0109, + "step": 106420 + }, + { + "epoch": 30.965958684899622, + "grad_norm": 0.1820676177740097, + "learning_rate": 1.0368290484153908e-06, + "loss": 0.0101, + "step": 106430 + }, + { + "epoch": 30.968868199010764, + "grad_norm": 0.12512712180614471, + "learning_rate": 1.0353194231935431e-06, + "loss": 0.0122, + "step": 106440 + }, + { + "epoch": 30.97177771312191, + "grad_norm": 0.18369877338409424, + "learning_rate": 1.0338108584970247e-06, + "loss": 0.0099, + "step": 106450 + }, + { + "epoch": 30.97468722723305, + "grad_norm": 0.09766913205385208, + "learning_rate": 1.0323033544404003e-06, + "loss": 0.0116, + "step": 106460 + }, + { + "epoch": 30.977596741344197, + "grad_norm": 0.0816168487071991, + "learning_rate": 1.0307969111381588e-06, + "loss": 0.0112, + "step": 106470 + }, + { + "epoch": 30.98050625545534, + "grad_norm": 0.09640775620937347, + "learning_rate": 1.0292915287047033e-06, + "loss": 0.01, + "step": 106480 + }, + { + "epoch": 30.98341576956648, + "grad_norm": 0.16401654481887817, + "learning_rate": 1.0277872072543576e-06, + "loss": 0.0129, + "step": 106490 + }, + { + "epoch": 30.986325283677626, + "grad_norm": 0.0927242636680603, + "learning_rate": 1.0262839469013653e-06, + "loss": 0.0089, + "step": 106500 + }, + { + "epoch": 30.989234797788768, + "grad_norm": 0.11998310685157776, + "learning_rate": 1.0247817477598882e-06, + "loss": 0.0112, + "step": 106510 + }, + { + "epoch": 30.992144311899914, + "grad_norm": 0.09637358039617538, + "learning_rate": 1.02328060994401e-06, + "loss": 0.0116, + "step": 106520 + }, + { + "epoch": 30.995053826011056, + "grad_norm": 0.12663063406944275, + "learning_rate": 1.0217805335677282e-06, + "loss": 0.0116, + "step": 106530 + }, + { + "epoch": 30.9979633401222, + "grad_norm": 0.10349169373512268, + "learning_rate": 1.0202815187449683e-06, + "loss": 0.0104, + "step": 106540 + }, + { + "epoch": 31.000872854233343, + "grad_norm": 0.08579937368631363, + "learning_rate": 1.0187835655895706e-06, + "loss": 0.0105, + "step": 106550 + }, + { + "epoch": 31.003782368344485, + "grad_norm": 0.11744187027215958, + "learning_rate": 1.017286674215291e-06, + "loss": 0.0096, + "step": 106560 + }, + { + "epoch": 31.00669188245563, + "grad_norm": 0.06384608149528503, + "learning_rate": 1.0157908447358117e-06, + "loss": 0.0093, + "step": 106570 + }, + { + "epoch": 31.009601396566772, + "grad_norm": 0.11971959471702576, + "learning_rate": 1.0142960772647285e-06, + "loss": 0.0084, + "step": 106580 + }, + { + "epoch": 31.012510910677918, + "grad_norm": 0.17587754130363464, + "learning_rate": 1.012802371915561e-06, + "loss": 0.0086, + "step": 106590 + }, + { + "epoch": 31.01542042478906, + "grad_norm": 0.07149980217218399, + "learning_rate": 1.0113097288017481e-06, + "loss": 0.0088, + "step": 106600 + }, + { + "epoch": 31.018329938900205, + "grad_norm": 0.07799102365970612, + "learning_rate": 1.0098181480366442e-06, + "loss": 0.0096, + "step": 106610 + }, + { + "epoch": 31.021239453011347, + "grad_norm": 0.10590581595897675, + "learning_rate": 1.008327629733522e-06, + "loss": 0.0094, + "step": 106620 + }, + { + "epoch": 31.02414896712249, + "grad_norm": 0.13652969896793365, + "learning_rate": 1.0068381740055792e-06, + "loss": 0.0104, + "step": 106630 + }, + { + "epoch": 31.027058481233635, + "grad_norm": 0.065923310816288, + "learning_rate": 1.005349780965934e-06, + "loss": 0.0114, + "step": 106640 + }, + { + "epoch": 31.029967995344776, + "grad_norm": 0.14813199639320374, + "learning_rate": 1.0038624507276123e-06, + "loss": 0.0124, + "step": 106650 + }, + { + "epoch": 31.032877509455922, + "grad_norm": 0.10884464532136917, + "learning_rate": 1.0023761834035706e-06, + "loss": 0.0107, + "step": 106660 + }, + { + "epoch": 31.035787023567064, + "grad_norm": 0.11700153350830078, + "learning_rate": 1.000890979106685e-06, + "loss": 0.0118, + "step": 106670 + }, + { + "epoch": 31.03869653767821, + "grad_norm": 0.12013102322816849, + "learning_rate": 9.99406837949744e-07, + "loss": 0.0104, + "step": 106680 + }, + { + "epoch": 31.04160605178935, + "grad_norm": 0.11013998836278915, + "learning_rate": 9.97923760045454e-07, + "loss": 0.0114, + "step": 106690 + }, + { + "epoch": 31.044515565900493, + "grad_norm": 0.10817331075668335, + "learning_rate": 9.964417455064495e-07, + "loss": 0.0116, + "step": 106700 + }, + { + "epoch": 31.04742508001164, + "grad_norm": 0.08787331730127335, + "learning_rate": 9.94960794445281e-07, + "loss": 0.0118, + "step": 106710 + }, + { + "epoch": 31.05033459412278, + "grad_norm": 0.0921345129609108, + "learning_rate": 9.93480906974415e-07, + "loss": 0.0113, + "step": 106720 + }, + { + "epoch": 31.053244108233926, + "grad_norm": 0.12044978141784668, + "learning_rate": 9.920020832062348e-07, + "loss": 0.0115, + "step": 106730 + }, + { + "epoch": 31.056153622345068, + "grad_norm": 0.09098994731903076, + "learning_rate": 9.905243232530559e-07, + "loss": 0.0109, + "step": 106740 + }, + { + "epoch": 31.059063136456214, + "grad_norm": 0.11655297875404358, + "learning_rate": 9.890476272270998e-07, + "loss": 0.0106, + "step": 106750 + }, + { + "epoch": 31.061972650567355, + "grad_norm": 0.09872806072235107, + "learning_rate": 9.875719952405104e-07, + "loss": 0.0101, + "step": 106760 + }, + { + "epoch": 31.064882164678497, + "grad_norm": 0.13221772015094757, + "learning_rate": 9.860974274053547e-07, + "loss": 0.0113, + "step": 106770 + }, + { + "epoch": 31.067791678789643, + "grad_norm": 0.08690030127763748, + "learning_rate": 9.846239238336163e-07, + "loss": 0.0111, + "step": 106780 + }, + { + "epoch": 31.070701192900785, + "grad_norm": 0.1728973686695099, + "learning_rate": 9.831514846371974e-07, + "loss": 0.0095, + "step": 106790 + }, + { + "epoch": 31.07361070701193, + "grad_norm": 0.0826890766620636, + "learning_rate": 9.816801099279204e-07, + "loss": 0.01, + "step": 106800 + }, + { + "epoch": 31.076520221123072, + "grad_norm": 0.08644115179777145, + "learning_rate": 9.80209799817524e-07, + "loss": 0.0099, + "step": 106810 + }, + { + "epoch": 31.079429735234214, + "grad_norm": 0.056658245623111725, + "learning_rate": 9.787405544176743e-07, + "loss": 0.0098, + "step": 106820 + }, + { + "epoch": 31.08233924934536, + "grad_norm": 0.08252207189798355, + "learning_rate": 9.772723738399448e-07, + "loss": 0.008, + "step": 106830 + }, + { + "epoch": 31.0852487634565, + "grad_norm": 0.16919498145580292, + "learning_rate": 9.75805258195835e-07, + "loss": 0.0104, + "step": 106840 + }, + { + "epoch": 31.088158277567647, + "grad_norm": 0.08135602623224258, + "learning_rate": 9.743392075967688e-07, + "loss": 0.012, + "step": 106850 + }, + { + "epoch": 31.09106779167879, + "grad_norm": 0.08779370039701462, + "learning_rate": 9.728742221540744e-07, + "loss": 0.0101, + "step": 106860 + }, + { + "epoch": 31.093977305789934, + "grad_norm": 0.13295549154281616, + "learning_rate": 9.714103019790137e-07, + "loss": 0.0105, + "step": 106870 + }, + { + "epoch": 31.096886819901076, + "grad_norm": 0.09053894877433777, + "learning_rate": 9.699474471827585e-07, + "loss": 0.0111, + "step": 106880 + }, + { + "epoch": 31.09979633401222, + "grad_norm": 0.058107681572437286, + "learning_rate": 9.68485657876408e-07, + "loss": 0.0108, + "step": 106890 + }, + { + "epoch": 31.102705848123364, + "grad_norm": 0.08478999882936478, + "learning_rate": 9.670249341709681e-07, + "loss": 0.0124, + "step": 106900 + }, + { + "epoch": 31.105615362234506, + "grad_norm": 0.10562200099229813, + "learning_rate": 9.65565276177377e-07, + "loss": 0.0119, + "step": 106910 + }, + { + "epoch": 31.10852487634565, + "grad_norm": 0.09630890190601349, + "learning_rate": 9.641066840064827e-07, + "loss": 0.0105, + "step": 106920 + }, + { + "epoch": 31.111434390456793, + "grad_norm": 0.146344855427742, + "learning_rate": 9.62649157769055e-07, + "loss": 0.0115, + "step": 106930 + }, + { + "epoch": 31.11434390456794, + "grad_norm": 0.1401563435792923, + "learning_rate": 9.6119269757579e-07, + "loss": 0.0093, + "step": 106940 + }, + { + "epoch": 31.11725341867908, + "grad_norm": 0.11713916808366776, + "learning_rate": 9.59737303537288e-07, + "loss": 0.0112, + "step": 106950 + }, + { + "epoch": 31.120162932790222, + "grad_norm": 0.10461767017841339, + "learning_rate": 9.582829757640804e-07, + "loss": 0.0091, + "step": 106960 + }, + { + "epoch": 31.123072446901368, + "grad_norm": 0.11929894983768463, + "learning_rate": 9.568297143666154e-07, + "loss": 0.0102, + "step": 106970 + }, + { + "epoch": 31.12598196101251, + "grad_norm": 0.16574062407016754, + "learning_rate": 9.553775194552566e-07, + "loss": 0.0108, + "step": 106980 + }, + { + "epoch": 31.128891475123655, + "grad_norm": 0.0890357717871666, + "learning_rate": 9.539263911402873e-07, + "loss": 0.0112, + "step": 106990 + }, + { + "epoch": 31.131800989234797, + "grad_norm": 0.08795329928398132, + "learning_rate": 9.524763295319127e-07, + "loss": 0.0094, + "step": 107000 + }, + { + "epoch": 31.134710503345943, + "grad_norm": 0.09672216325998306, + "learning_rate": 9.510273347402565e-07, + "loss": 0.0132, + "step": 107010 + }, + { + "epoch": 31.137620017457085, + "grad_norm": 0.12338461726903915, + "learning_rate": 9.495794068753588e-07, + "loss": 0.0108, + "step": 107020 + }, + { + "epoch": 31.140529531568227, + "grad_norm": 0.07507596164941788, + "learning_rate": 9.481325460471769e-07, + "loss": 0.0094, + "step": 107030 + }, + { + "epoch": 31.143439045679372, + "grad_norm": 0.06890665739774704, + "learning_rate": 9.466867523655992e-07, + "loss": 0.0124, + "step": 107040 + }, + { + "epoch": 31.146348559790514, + "grad_norm": 0.1802268773317337, + "learning_rate": 9.452420259404166e-07, + "loss": 0.0107, + "step": 107050 + }, + { + "epoch": 31.14925807390166, + "grad_norm": 0.1503475159406662, + "learning_rate": 9.437983668813477e-07, + "loss": 0.0097, + "step": 107060 + }, + { + "epoch": 31.1521675880128, + "grad_norm": 0.07442568242549896, + "learning_rate": 9.4235577529803e-07, + "loss": 0.0104, + "step": 107070 + }, + { + "epoch": 31.155077102123947, + "grad_norm": 0.2562362253665924, + "learning_rate": 9.409142513000224e-07, + "loss": 0.0097, + "step": 107080 + }, + { + "epoch": 31.15798661623509, + "grad_norm": 0.11122303456068039, + "learning_rate": 9.394737949967941e-07, + "loss": 0.0122, + "step": 107090 + }, + { + "epoch": 31.16089613034623, + "grad_norm": 0.11476790904998779, + "learning_rate": 9.380344064977376e-07, + "loss": 0.0111, + "step": 107100 + }, + { + "epoch": 31.163805644457376, + "grad_norm": 0.08415137976408005, + "learning_rate": 9.36596085912167e-07, + "loss": 0.0085, + "step": 107110 + }, + { + "epoch": 31.16671515856852, + "grad_norm": 0.11105503886938095, + "learning_rate": 9.351588333493149e-07, + "loss": 0.0088, + "step": 107120 + }, + { + "epoch": 31.169624672679664, + "grad_norm": 0.08115503937005997, + "learning_rate": 9.337226489183275e-07, + "loss": 0.0089, + "step": 107130 + }, + { + "epoch": 31.172534186790806, + "grad_norm": 0.0769832581281662, + "learning_rate": 9.322875327282759e-07, + "loss": 0.0115, + "step": 107140 + }, + { + "epoch": 31.17544370090195, + "grad_norm": 0.207894429564476, + "learning_rate": 9.308534848881495e-07, + "loss": 0.0111, + "step": 107150 + }, + { + "epoch": 31.178353215013093, + "grad_norm": 0.08494050800800323, + "learning_rate": 9.294205055068511e-07, + "loss": 0.0103, + "step": 107160 + }, + { + "epoch": 31.181262729124235, + "grad_norm": 0.0958627387881279, + "learning_rate": 9.279885946932088e-07, + "loss": 0.0098, + "step": 107170 + }, + { + "epoch": 31.18417224323538, + "grad_norm": 0.07292490452528, + "learning_rate": 9.265577525559637e-07, + "loss": 0.0103, + "step": 107180 + }, + { + "epoch": 31.187081757346522, + "grad_norm": 0.10469215363264084, + "learning_rate": 9.251279792037843e-07, + "loss": 0.0091, + "step": 107190 + }, + { + "epoch": 31.189991271457668, + "grad_norm": 0.1063617616891861, + "learning_rate": 9.23699274745245e-07, + "loss": 0.0113, + "step": 107200 + }, + { + "epoch": 31.19290078556881, + "grad_norm": 0.10038509219884872, + "learning_rate": 9.222716392888525e-07, + "loss": 0.0096, + "step": 107210 + }, + { + "epoch": 31.195810299679952, + "grad_norm": 0.1481461077928543, + "learning_rate": 9.208450729430217e-07, + "loss": 0.0134, + "step": 107220 + }, + { + "epoch": 31.198719813791097, + "grad_norm": 0.08617675304412842, + "learning_rate": 9.194195758160944e-07, + "loss": 0.0109, + "step": 107230 + }, + { + "epoch": 31.20162932790224, + "grad_norm": 0.07041136920452118, + "learning_rate": 9.179951480163273e-07, + "loss": 0.0094, + "step": 107240 + }, + { + "epoch": 31.204538842013385, + "grad_norm": 0.0943809300661087, + "learning_rate": 9.165717896518938e-07, + "loss": 0.0091, + "step": 107250 + }, + { + "epoch": 31.207448356124527, + "grad_norm": 0.1798590123653412, + "learning_rate": 9.151495008308941e-07, + "loss": 0.011, + "step": 107260 + }, + { + "epoch": 31.210357870235672, + "grad_norm": 0.12950661778450012, + "learning_rate": 9.137282816613335e-07, + "loss": 0.0117, + "step": 107270 + }, + { + "epoch": 31.213267384346814, + "grad_norm": 0.10606164485216141, + "learning_rate": 9.123081322511523e-07, + "loss": 0.0093, + "step": 107280 + }, + { + "epoch": 31.216176898457956, + "grad_norm": 0.0800674632191658, + "learning_rate": 9.108890527081942e-07, + "loss": 0.0097, + "step": 107290 + }, + { + "epoch": 31.2190864125691, + "grad_norm": 0.0868753120303154, + "learning_rate": 9.094710431402331e-07, + "loss": 0.0094, + "step": 107300 + }, + { + "epoch": 31.221995926680243, + "grad_norm": 0.1068207398056984, + "learning_rate": 9.080541036549594e-07, + "loss": 0.0118, + "step": 107310 + }, + { + "epoch": 31.22490544079139, + "grad_norm": 0.10670715570449829, + "learning_rate": 9.066382343599789e-07, + "loss": 0.011, + "step": 107320 + }, + { + "epoch": 31.22781495490253, + "grad_norm": 0.12495718896389008, + "learning_rate": 9.052234353628103e-07, + "loss": 0.0122, + "step": 107330 + }, + { + "epoch": 31.230724469013676, + "grad_norm": 0.15663863718509674, + "learning_rate": 9.038097067709094e-07, + "loss": 0.0092, + "step": 107340 + }, + { + "epoch": 31.233633983124818, + "grad_norm": 0.08742251992225647, + "learning_rate": 9.023970486916339e-07, + "loss": 0.0115, + "step": 107350 + }, + { + "epoch": 31.23654349723596, + "grad_norm": 0.11078950762748718, + "learning_rate": 9.009854612322643e-07, + "loss": 0.0111, + "step": 107360 + }, + { + "epoch": 31.239453011347106, + "grad_norm": 0.12824976444244385, + "learning_rate": 8.995749445000051e-07, + "loss": 0.011, + "step": 107370 + }, + { + "epoch": 31.242362525458248, + "grad_norm": 0.10257257521152496, + "learning_rate": 8.981654986019739e-07, + "loss": 0.0092, + "step": 107380 + }, + { + "epoch": 31.245272039569393, + "grad_norm": 0.12422682344913483, + "learning_rate": 8.967571236452116e-07, + "loss": 0.0098, + "step": 107390 + }, + { + "epoch": 31.248181553680535, + "grad_norm": 0.13038231432437897, + "learning_rate": 8.953498197366677e-07, + "loss": 0.0093, + "step": 107400 + }, + { + "epoch": 31.25109106779168, + "grad_norm": 0.11805083602666855, + "learning_rate": 8.939435869832235e-07, + "loss": 0.0087, + "step": 107410 + }, + { + "epoch": 31.254000581902822, + "grad_norm": 0.12590573728084564, + "learning_rate": 8.925384254916752e-07, + "loss": 0.0135, + "step": 107420 + }, + { + "epoch": 31.256910096013964, + "grad_norm": 0.09681171178817749, + "learning_rate": 8.91134335368729e-07, + "loss": 0.0095, + "step": 107430 + }, + { + "epoch": 31.25981961012511, + "grad_norm": 0.08039841800928116, + "learning_rate": 8.897313167210197e-07, + "loss": 0.0097, + "step": 107440 + }, + { + "epoch": 31.26272912423625, + "grad_norm": 0.08205942809581757, + "learning_rate": 8.883293696551003e-07, + "loss": 0.0084, + "step": 107450 + }, + { + "epoch": 31.265638638347397, + "grad_norm": 0.13742157816886902, + "learning_rate": 8.869284942774359e-07, + "loss": 0.0111, + "step": 107460 + }, + { + "epoch": 31.26854815245854, + "grad_norm": 0.07365301996469498, + "learning_rate": 8.85528690694411e-07, + "loss": 0.0107, + "step": 107470 + }, + { + "epoch": 31.27145766656968, + "grad_norm": 0.11580245941877365, + "learning_rate": 8.841299590123358e-07, + "loss": 0.0122, + "step": 107480 + }, + { + "epoch": 31.274367180680827, + "grad_norm": 0.1585770696401596, + "learning_rate": 8.827322993374354e-07, + "loss": 0.0095, + "step": 107490 + }, + { + "epoch": 31.27727669479197, + "grad_norm": 0.1696770042181015, + "learning_rate": 8.813357117758497e-07, + "loss": 0.0115, + "step": 107500 + }, + { + "epoch": 31.280186208903114, + "grad_norm": 0.13954579830169678, + "learning_rate": 8.799401964336423e-07, + "loss": 0.009, + "step": 107510 + }, + { + "epoch": 31.283095723014256, + "grad_norm": 0.13162095844745636, + "learning_rate": 8.785457534167918e-07, + "loss": 0.01, + "step": 107520 + }, + { + "epoch": 31.2860052371254, + "grad_norm": 0.13236530125141144, + "learning_rate": 8.771523828311967e-07, + "loss": 0.009, + "step": 107530 + }, + { + "epoch": 31.288914751236543, + "grad_norm": 0.18253971636295319, + "learning_rate": 8.757600847826775e-07, + "loss": 0.0102, + "step": 107540 + }, + { + "epoch": 31.291824265347685, + "grad_norm": 0.09206025302410126, + "learning_rate": 8.743688593769661e-07, + "loss": 0.0099, + "step": 107550 + }, + { + "epoch": 31.29473377945883, + "grad_norm": 0.10042881220579147, + "learning_rate": 8.729787067197215e-07, + "loss": 0.0107, + "step": 107560 + }, + { + "epoch": 31.297643293569973, + "grad_norm": 0.10960332304239273, + "learning_rate": 8.715896269165124e-07, + "loss": 0.0094, + "step": 107570 + }, + { + "epoch": 31.300552807681118, + "grad_norm": 0.14538423717021942, + "learning_rate": 8.70201620072833e-07, + "loss": 0.0121, + "step": 107580 + }, + { + "epoch": 31.30346232179226, + "grad_norm": 0.0609954297542572, + "learning_rate": 8.688146862940887e-07, + "loss": 0.0097, + "step": 107590 + }, + { + "epoch": 31.306371835903406, + "grad_norm": 0.08499401062726974, + "learning_rate": 8.674288256856139e-07, + "loss": 0.0116, + "step": 107600 + }, + { + "epoch": 31.309281350014547, + "grad_norm": 0.07817826420068741, + "learning_rate": 8.660440383526525e-07, + "loss": 0.011, + "step": 107610 + }, + { + "epoch": 31.31219086412569, + "grad_norm": 0.05835286155343056, + "learning_rate": 8.646603244003721e-07, + "loss": 0.0114, + "step": 107620 + }, + { + "epoch": 31.315100378236835, + "grad_norm": 0.12541556358337402, + "learning_rate": 8.632776839338502e-07, + "loss": 0.0101, + "step": 107630 + }, + { + "epoch": 31.318009892347977, + "grad_norm": 0.11655835807323456, + "learning_rate": 8.61896117058098e-07, + "loss": 0.0109, + "step": 107640 + }, + { + "epoch": 31.320919406459122, + "grad_norm": 0.09512688219547272, + "learning_rate": 8.605156238780315e-07, + "loss": 0.0113, + "step": 107650 + }, + { + "epoch": 31.323828920570264, + "grad_norm": 0.08537497371435165, + "learning_rate": 8.591362044984902e-07, + "loss": 0.0105, + "step": 107660 + }, + { + "epoch": 31.32673843468141, + "grad_norm": 0.24676743149757385, + "learning_rate": 8.577578590242319e-07, + "loss": 0.0098, + "step": 107670 + }, + { + "epoch": 31.32964794879255, + "grad_norm": 0.08566449582576752, + "learning_rate": 8.563805875599362e-07, + "loss": 0.0092, + "step": 107680 + }, + { + "epoch": 31.332557462903694, + "grad_norm": 0.11563721299171448, + "learning_rate": 8.550043902101945e-07, + "loss": 0.0098, + "step": 107690 + }, + { + "epoch": 31.33546697701484, + "grad_norm": 0.10295818001031876, + "learning_rate": 8.536292670795181e-07, + "loss": 0.008, + "step": 107700 + }, + { + "epoch": 31.33837649112598, + "grad_norm": 0.09013421088457108, + "learning_rate": 8.52255218272342e-07, + "loss": 0.0129, + "step": 107710 + }, + { + "epoch": 31.341286005237126, + "grad_norm": 0.11347890645265579, + "learning_rate": 8.508822438930175e-07, + "loss": 0.0132, + "step": 107720 + }, + { + "epoch": 31.34419551934827, + "grad_norm": 0.1449856460094452, + "learning_rate": 8.495103440458097e-07, + "loss": 0.0104, + "step": 107730 + }, + { + "epoch": 31.347105033459414, + "grad_norm": 0.14865034818649292, + "learning_rate": 8.481395188349017e-07, + "loss": 0.0093, + "step": 107740 + }, + { + "epoch": 31.350014547570556, + "grad_norm": 0.08852461725473404, + "learning_rate": 8.467697683644071e-07, + "loss": 0.0117, + "step": 107750 + }, + { + "epoch": 31.352924061681698, + "grad_norm": 0.08102378994226456, + "learning_rate": 8.454010927383459e-07, + "loss": 0.0096, + "step": 107760 + }, + { + "epoch": 31.355833575792843, + "grad_norm": 0.13631099462509155, + "learning_rate": 8.440334920606568e-07, + "loss": 0.009, + "step": 107770 + }, + { + "epoch": 31.358743089903985, + "grad_norm": 0.08329478651285172, + "learning_rate": 8.426669664352032e-07, + "loss": 0.0097, + "step": 107780 + }, + { + "epoch": 31.36165260401513, + "grad_norm": 0.10765507072210312, + "learning_rate": 8.413015159657656e-07, + "loss": 0.0113, + "step": 107790 + }, + { + "epoch": 31.364562118126273, + "grad_norm": 0.10821696370840073, + "learning_rate": 8.399371407560358e-07, + "loss": 0.0102, + "step": 107800 + }, + { + "epoch": 31.367471632237418, + "grad_norm": 0.12507161498069763, + "learning_rate": 8.385738409096344e-07, + "loss": 0.0089, + "step": 107810 + }, + { + "epoch": 31.37038114634856, + "grad_norm": 0.16274262964725494, + "learning_rate": 8.37211616530092e-07, + "loss": 0.0091, + "step": 107820 + }, + { + "epoch": 31.373290660459702, + "grad_norm": 0.12736724317073822, + "learning_rate": 8.358504677208606e-07, + "loss": 0.0112, + "step": 107830 + }, + { + "epoch": 31.376200174570847, + "grad_norm": 0.09073193371295929, + "learning_rate": 8.344903945853094e-07, + "loss": 0.0104, + "step": 107840 + }, + { + "epoch": 31.37910968868199, + "grad_norm": 0.1007775366306305, + "learning_rate": 8.33131397226729e-07, + "loss": 0.0102, + "step": 107850 + }, + { + "epoch": 31.382019202793135, + "grad_norm": 0.11953368037939072, + "learning_rate": 8.317734757483287e-07, + "loss": 0.012, + "step": 107860 + }, + { + "epoch": 31.384928716904277, + "grad_norm": 0.12205745279788971, + "learning_rate": 8.304166302532274e-07, + "loss": 0.0112, + "step": 107870 + }, + { + "epoch": 31.38783823101542, + "grad_norm": 0.08809947222471237, + "learning_rate": 8.290608608444744e-07, + "loss": 0.0107, + "step": 107880 + }, + { + "epoch": 31.390747745126564, + "grad_norm": 0.09555160254240036, + "learning_rate": 8.277061676250275e-07, + "loss": 0.0106, + "step": 107890 + }, + { + "epoch": 31.393657259237706, + "grad_norm": 0.11543082445859909, + "learning_rate": 8.263525506977659e-07, + "loss": 0.0106, + "step": 107900 + }, + { + "epoch": 31.39656677334885, + "grad_norm": 0.11570753157138824, + "learning_rate": 8.250000101654942e-07, + "loss": 0.01, + "step": 107910 + }, + { + "epoch": 31.399476287459994, + "grad_norm": 0.16525398194789886, + "learning_rate": 8.236485461309251e-07, + "loss": 0.0117, + "step": 107920 + }, + { + "epoch": 31.40238580157114, + "grad_norm": 0.19606074690818787, + "learning_rate": 8.222981586966899e-07, + "loss": 0.0104, + "step": 107930 + }, + { + "epoch": 31.40529531568228, + "grad_norm": 0.11574406176805496, + "learning_rate": 8.209488479653465e-07, + "loss": 0.0096, + "step": 107940 + }, + { + "epoch": 31.408204829793423, + "grad_norm": 0.08458445966243744, + "learning_rate": 8.196006140393648e-07, + "loss": 0.0111, + "step": 107950 + }, + { + "epoch": 31.41111434390457, + "grad_norm": 0.07316742837429047, + "learning_rate": 8.182534570211325e-07, + "loss": 0.0091, + "step": 107960 + }, + { + "epoch": 31.41402385801571, + "grad_norm": 0.11184930801391602, + "learning_rate": 8.169073770129582e-07, + "loss": 0.0098, + "step": 107970 + }, + { + "epoch": 31.416933372126856, + "grad_norm": 0.09510987997055054, + "learning_rate": 8.155623741170715e-07, + "loss": 0.012, + "step": 107980 + }, + { + "epoch": 31.419842886237998, + "grad_norm": 0.10109162330627441, + "learning_rate": 8.142184484356141e-07, + "loss": 0.0121, + "step": 107990 + }, + { + "epoch": 31.422752400349143, + "grad_norm": 0.12058918178081512, + "learning_rate": 8.128756000706427e-07, + "loss": 0.0103, + "step": 108000 + }, + { + "epoch": 31.425661914460285, + "grad_norm": 0.07423493266105652, + "learning_rate": 8.115338291241442e-07, + "loss": 0.0111, + "step": 108010 + }, + { + "epoch": 31.428571428571427, + "grad_norm": 0.2692904770374298, + "learning_rate": 8.101931356980185e-07, + "loss": 0.0104, + "step": 108020 + }, + { + "epoch": 31.431480942682573, + "grad_norm": 0.10104429721832275, + "learning_rate": 8.088535198940794e-07, + "loss": 0.0097, + "step": 108030 + }, + { + "epoch": 31.434390456793714, + "grad_norm": 0.1152147427201271, + "learning_rate": 8.07514981814057e-07, + "loss": 0.0095, + "step": 108040 + }, + { + "epoch": 31.43729997090486, + "grad_norm": 0.09975415468215942, + "learning_rate": 8.061775215596135e-07, + "loss": 0.0132, + "step": 108050 + }, + { + "epoch": 31.440209485016002, + "grad_norm": 0.12597370147705078, + "learning_rate": 8.048411392323157e-07, + "loss": 0.0107, + "step": 108060 + }, + { + "epoch": 31.443118999127147, + "grad_norm": 0.07578849792480469, + "learning_rate": 8.035058349336527e-07, + "loss": 0.0112, + "step": 108070 + }, + { + "epoch": 31.44602851323829, + "grad_norm": 0.07084766030311584, + "learning_rate": 8.021716087650332e-07, + "loss": 0.0101, + "step": 108080 + }, + { + "epoch": 31.44893802734943, + "grad_norm": 0.14158760011196136, + "learning_rate": 8.008384608277831e-07, + "loss": 0.0101, + "step": 108090 + }, + { + "epoch": 31.451847541460577, + "grad_norm": 0.07354629784822464, + "learning_rate": 7.99506391223146e-07, + "loss": 0.0103, + "step": 108100 + }, + { + "epoch": 31.45475705557172, + "grad_norm": 0.059892114251852036, + "learning_rate": 7.981754000522812e-07, + "loss": 0.009, + "step": 108110 + }, + { + "epoch": 31.457666569682864, + "grad_norm": 0.1404399573802948, + "learning_rate": 7.968454874162711e-07, + "loss": 0.0104, + "step": 108120 + }, + { + "epoch": 31.460576083794006, + "grad_norm": 0.10901843756437302, + "learning_rate": 7.955166534161168e-07, + "loss": 0.0086, + "step": 108130 + }, + { + "epoch": 31.46348559790515, + "grad_norm": 0.11987591534852982, + "learning_rate": 7.941888981527274e-07, + "loss": 0.0092, + "step": 108140 + }, + { + "epoch": 31.466395112016293, + "grad_norm": 0.11374425143003464, + "learning_rate": 7.928622217269421e-07, + "loss": 0.0107, + "step": 108150 + }, + { + "epoch": 31.469304626127435, + "grad_norm": 0.09781233221292496, + "learning_rate": 7.915366242395139e-07, + "loss": 0.0102, + "step": 108160 + }, + { + "epoch": 31.47221414023858, + "grad_norm": 0.17467987537384033, + "learning_rate": 7.902121057911088e-07, + "loss": 0.0117, + "step": 108170 + }, + { + "epoch": 31.475123654349723, + "grad_norm": 0.121156245470047, + "learning_rate": 7.888886664823214e-07, + "loss": 0.0097, + "step": 108180 + }, + { + "epoch": 31.47803316846087, + "grad_norm": 0.13338503241539001, + "learning_rate": 7.875663064136512e-07, + "loss": 0.01, + "step": 108190 + }, + { + "epoch": 31.48094268257201, + "grad_norm": 0.11696972697973251, + "learning_rate": 7.862450256855281e-07, + "loss": 0.01, + "step": 108200 + }, + { + "epoch": 31.483852196683152, + "grad_norm": 0.11656399071216583, + "learning_rate": 7.849248243982915e-07, + "loss": 0.0101, + "step": 108210 + }, + { + "epoch": 31.486761710794298, + "grad_norm": 0.1254146546125412, + "learning_rate": 7.836057026522048e-07, + "loss": 0.012, + "step": 108220 + }, + { + "epoch": 31.48967122490544, + "grad_norm": 0.15225361287593842, + "learning_rate": 7.822876605474427e-07, + "loss": 0.0102, + "step": 108230 + }, + { + "epoch": 31.492580739016585, + "grad_norm": 0.11466509103775024, + "learning_rate": 7.809706981841036e-07, + "loss": 0.0102, + "step": 108240 + }, + { + "epoch": 31.495490253127727, + "grad_norm": 0.08512228727340698, + "learning_rate": 7.796548156622057e-07, + "loss": 0.0107, + "step": 108250 + }, + { + "epoch": 31.498399767238872, + "grad_norm": 0.0906011313199997, + "learning_rate": 7.783400130816759e-07, + "loss": 0.0108, + "step": 108260 + }, + { + "epoch": 31.501309281350014, + "grad_norm": 0.1115039587020874, + "learning_rate": 7.770262905423659e-07, + "loss": 0.0098, + "step": 108270 + }, + { + "epoch": 31.504218795461156, + "grad_norm": 0.1675877571105957, + "learning_rate": 7.757136481440491e-07, + "loss": 0.0098, + "step": 108280 + }, + { + "epoch": 31.507128309572302, + "grad_norm": 0.08302187919616699, + "learning_rate": 7.744020859864092e-07, + "loss": 0.0096, + "step": 108290 + }, + { + "epoch": 31.510037823683444, + "grad_norm": 0.09591837972402573, + "learning_rate": 7.730916041690467e-07, + "loss": 0.0117, + "step": 108300 + }, + { + "epoch": 31.51294733779459, + "grad_norm": 0.13310286402702332, + "learning_rate": 7.717822027914883e-07, + "loss": 0.0129, + "step": 108310 + }, + { + "epoch": 31.51585685190573, + "grad_norm": 0.1580563634634018, + "learning_rate": 7.704738819531764e-07, + "loss": 0.0089, + "step": 108320 + }, + { + "epoch": 31.518766366016877, + "grad_norm": 0.10221477597951889, + "learning_rate": 7.691666417534648e-07, + "loss": 0.0102, + "step": 108330 + }, + { + "epoch": 31.52167588012802, + "grad_norm": 0.18526118993759155, + "learning_rate": 7.678604822916274e-07, + "loss": 0.0101, + "step": 108340 + }, + { + "epoch": 31.52458539423916, + "grad_norm": 0.08451761305332184, + "learning_rate": 7.665554036668682e-07, + "loss": 0.0101, + "step": 108350 + }, + { + "epoch": 31.527494908350306, + "grad_norm": 0.12834417819976807, + "learning_rate": 7.652514059782911e-07, + "loss": 0.011, + "step": 108360 + }, + { + "epoch": 31.530404422461448, + "grad_norm": 0.07597735524177551, + "learning_rate": 7.639484893249288e-07, + "loss": 0.01, + "step": 108370 + }, + { + "epoch": 31.533313936572593, + "grad_norm": 0.09491363912820816, + "learning_rate": 7.626466538057286e-07, + "loss": 0.0102, + "step": 108380 + }, + { + "epoch": 31.536223450683735, + "grad_norm": 0.15173976123332977, + "learning_rate": 7.613458995195582e-07, + "loss": 0.0098, + "step": 108390 + }, + { + "epoch": 31.53913296479488, + "grad_norm": 0.1389288306236267, + "learning_rate": 7.600462265651986e-07, + "loss": 0.0116, + "step": 108400 + }, + { + "epoch": 31.542042478906023, + "grad_norm": 0.10672543942928314, + "learning_rate": 7.587476350413525e-07, + "loss": 0.0111, + "step": 108410 + }, + { + "epoch": 31.544951993017165, + "grad_norm": 0.10481435060501099, + "learning_rate": 7.574501250466376e-07, + "loss": 0.0099, + "step": 108420 + }, + { + "epoch": 31.54786150712831, + "grad_norm": 0.09490357339382172, + "learning_rate": 7.561536966795968e-07, + "loss": 0.0086, + "step": 108430 + }, + { + "epoch": 31.550771021239452, + "grad_norm": 0.0860191285610199, + "learning_rate": 7.548583500386779e-07, + "loss": 0.0099, + "step": 108440 + }, + { + "epoch": 31.553680535350598, + "grad_norm": 0.10451877117156982, + "learning_rate": 7.535640852222591e-07, + "loss": 0.0124, + "step": 108450 + }, + { + "epoch": 31.55659004946174, + "grad_norm": 0.14720484614372253, + "learning_rate": 7.522709023286317e-07, + "loss": 0.0107, + "step": 108460 + }, + { + "epoch": 31.559499563572885, + "grad_norm": 0.0870155617594719, + "learning_rate": 7.509788014559988e-07, + "loss": 0.0113, + "step": 108470 + }, + { + "epoch": 31.562409077684027, + "grad_norm": 0.06232136860489845, + "learning_rate": 7.496877827024934e-07, + "loss": 0.0083, + "step": 108480 + }, + { + "epoch": 31.56531859179517, + "grad_norm": 0.1389746516942978, + "learning_rate": 7.483978461661556e-07, + "loss": 0.0135, + "step": 108490 + }, + { + "epoch": 31.568228105906314, + "grad_norm": 0.1309206336736679, + "learning_rate": 7.471089919449503e-07, + "loss": 0.01, + "step": 108500 + }, + { + "epoch": 31.571137620017456, + "grad_norm": 0.1809508502483368, + "learning_rate": 7.458212201367559e-07, + "loss": 0.0113, + "step": 108510 + }, + { + "epoch": 31.5740471341286, + "grad_norm": 0.10127103328704834, + "learning_rate": 7.445345308393708e-07, + "loss": 0.0103, + "step": 108520 + }, + { + "epoch": 31.576956648239744, + "grad_norm": 0.1478739082813263, + "learning_rate": 7.432489241505086e-07, + "loss": 0.0124, + "step": 108530 + }, + { + "epoch": 31.57986616235089, + "grad_norm": 0.1369992345571518, + "learning_rate": 7.419644001678027e-07, + "loss": 0.0096, + "step": 108540 + }, + { + "epoch": 31.58277567646203, + "grad_norm": 0.0946832075715065, + "learning_rate": 7.406809589888103e-07, + "loss": 0.0114, + "step": 108550 + }, + { + "epoch": 31.585685190573173, + "grad_norm": 0.11892380565404892, + "learning_rate": 7.393986007109915e-07, + "loss": 0.0108, + "step": 108560 + }, + { + "epoch": 31.58859470468432, + "grad_norm": 0.11446026712656021, + "learning_rate": 7.381173254317402e-07, + "loss": 0.0105, + "step": 108570 + }, + { + "epoch": 31.59150421879546, + "grad_norm": 0.13679124414920807, + "learning_rate": 7.36837133248357e-07, + "loss": 0.0102, + "step": 108580 + }, + { + "epoch": 31.594413732906606, + "grad_norm": 0.12235908210277557, + "learning_rate": 7.355580242580656e-07, + "loss": 0.0104, + "step": 108590 + }, + { + "epoch": 31.597323247017748, + "grad_norm": 0.10380154848098755, + "learning_rate": 7.342799985580035e-07, + "loss": 0.0118, + "step": 108600 + }, + { + "epoch": 31.60023276112889, + "grad_norm": 0.13381414115428925, + "learning_rate": 7.330030562452295e-07, + "loss": 0.0109, + "step": 108610 + }, + { + "epoch": 31.603142275240035, + "grad_norm": 0.07168683409690857, + "learning_rate": 7.317271974167211e-07, + "loss": 0.0095, + "step": 108620 + }, + { + "epoch": 31.606051789351177, + "grad_norm": 0.1474916934967041, + "learning_rate": 7.304524221693709e-07, + "loss": 0.0094, + "step": 108630 + }, + { + "epoch": 31.608961303462323, + "grad_norm": 0.08947078138589859, + "learning_rate": 7.29178730599983e-07, + "loss": 0.0102, + "step": 108640 + }, + { + "epoch": 31.611870817573465, + "grad_norm": 0.12340953201055527, + "learning_rate": 7.279061228052952e-07, + "loss": 0.0103, + "step": 108650 + }, + { + "epoch": 31.61478033168461, + "grad_norm": 0.10539325326681137, + "learning_rate": 7.266345988819518e-07, + "loss": 0.011, + "step": 108660 + }, + { + "epoch": 31.617689845795752, + "grad_norm": 0.06238440424203873, + "learning_rate": 7.253641589265103e-07, + "loss": 0.011, + "step": 108670 + }, + { + "epoch": 31.620599359906894, + "grad_norm": 0.11521542817354202, + "learning_rate": 7.240948030354572e-07, + "loss": 0.0081, + "step": 108680 + }, + { + "epoch": 31.62350887401804, + "grad_norm": 0.10123081505298615, + "learning_rate": 7.228265313051935e-07, + "loss": 0.01, + "step": 108690 + }, + { + "epoch": 31.62641838812918, + "grad_norm": 0.10136068612337112, + "learning_rate": 7.215593438320339e-07, + "loss": 0.011, + "step": 108700 + }, + { + "epoch": 31.629327902240327, + "grad_norm": 0.10492564737796783, + "learning_rate": 7.202932407122098e-07, + "loss": 0.0084, + "step": 108710 + }, + { + "epoch": 31.63223741635147, + "grad_norm": 0.12275492399930954, + "learning_rate": 7.190282220418775e-07, + "loss": 0.0101, + "step": 108720 + }, + { + "epoch": 31.635146930462614, + "grad_norm": 0.09442141652107239, + "learning_rate": 7.177642879171053e-07, + "loss": 0.0122, + "step": 108730 + }, + { + "epoch": 31.638056444573756, + "grad_norm": 0.09862438589334488, + "learning_rate": 7.165014384338814e-07, + "loss": 0.0103, + "step": 108740 + }, + { + "epoch": 31.640965958684898, + "grad_norm": 0.10215488076210022, + "learning_rate": 7.152396736881089e-07, + "loss": 0.01, + "step": 108750 + }, + { + "epoch": 31.643875472796044, + "grad_norm": 0.12798267602920532, + "learning_rate": 7.139789937756147e-07, + "loss": 0.0098, + "step": 108760 + }, + { + "epoch": 31.646784986907186, + "grad_norm": 0.10439151525497437, + "learning_rate": 7.127193987921354e-07, + "loss": 0.0115, + "step": 108770 + }, + { + "epoch": 31.64969450101833, + "grad_norm": 0.10075860470533371, + "learning_rate": 7.114608888333279e-07, + "loss": 0.009, + "step": 108780 + }, + { + "epoch": 31.652604015129473, + "grad_norm": 0.11163794994354248, + "learning_rate": 7.102034639947708e-07, + "loss": 0.0109, + "step": 108790 + }, + { + "epoch": 31.65551352924062, + "grad_norm": 0.11293158680200577, + "learning_rate": 7.089471243719575e-07, + "loss": 0.0099, + "step": 108800 + }, + { + "epoch": 31.65842304335176, + "grad_norm": 0.10602504760026932, + "learning_rate": 7.076918700602953e-07, + "loss": 0.0118, + "step": 108810 + }, + { + "epoch": 31.661332557462902, + "grad_norm": 0.29136061668395996, + "learning_rate": 7.06437701155116e-07, + "loss": 0.0099, + "step": 108820 + }, + { + "epoch": 31.664242071574048, + "grad_norm": 0.07644382864236832, + "learning_rate": 7.051846177516635e-07, + "loss": 0.0097, + "step": 108830 + }, + { + "epoch": 31.66715158568519, + "grad_norm": 0.13200487196445465, + "learning_rate": 7.039326199451002e-07, + "loss": 0.0102, + "step": 108840 + }, + { + "epoch": 31.670061099796335, + "grad_norm": 0.10131537169218063, + "learning_rate": 7.026817078305115e-07, + "loss": 0.0093, + "step": 108850 + }, + { + "epoch": 31.672970613907477, + "grad_norm": 0.23050780594348907, + "learning_rate": 7.014318815028914e-07, + "loss": 0.0113, + "step": 108860 + }, + { + "epoch": 31.67588012801862, + "grad_norm": 0.11968255788087845, + "learning_rate": 7.001831410571591e-07, + "loss": 0.0099, + "step": 108870 + }, + { + "epoch": 31.678789642129765, + "grad_norm": 0.07870137691497803, + "learning_rate": 6.989354865881436e-07, + "loss": 0.0108, + "step": 108880 + }, + { + "epoch": 31.681699156240906, + "grad_norm": 0.12956269085407257, + "learning_rate": 6.976889181906027e-07, + "loss": 0.0098, + "step": 108890 + }, + { + "epoch": 31.684608670352052, + "grad_norm": 0.1083388552069664, + "learning_rate": 6.964434359591987e-07, + "loss": 0.0109, + "step": 108900 + }, + { + "epoch": 31.687518184463194, + "grad_norm": 0.1051262840628624, + "learning_rate": 6.95199039988521e-07, + "loss": 0.0102, + "step": 108910 + }, + { + "epoch": 31.69042769857434, + "grad_norm": 0.11298222094774246, + "learning_rate": 6.939557303730743e-07, + "loss": 0.0091, + "step": 108920 + }, + { + "epoch": 31.69333721268548, + "grad_norm": 0.14410999417304993, + "learning_rate": 6.927135072072777e-07, + "loss": 0.0124, + "step": 108930 + }, + { + "epoch": 31.696246726796623, + "grad_norm": 0.09204185009002686, + "learning_rate": 6.914723705854692e-07, + "loss": 0.0098, + "step": 108940 + }, + { + "epoch": 31.69915624090777, + "grad_norm": 0.09973140060901642, + "learning_rate": 6.902323206019051e-07, + "loss": 0.0079, + "step": 108950 + }, + { + "epoch": 31.70206575501891, + "grad_norm": 0.06893222033977509, + "learning_rate": 6.889933573507618e-07, + "loss": 0.0107, + "step": 108960 + }, + { + "epoch": 31.704975269130056, + "grad_norm": 0.08530843257904053, + "learning_rate": 6.877554809261289e-07, + "loss": 0.0102, + "step": 108970 + }, + { + "epoch": 31.707884783241198, + "grad_norm": 0.06420441716909409, + "learning_rate": 6.865186914220128e-07, + "loss": 0.0092, + "step": 108980 + }, + { + "epoch": 31.710794297352344, + "grad_norm": 0.17383909225463867, + "learning_rate": 6.852829889323436e-07, + "loss": 0.0104, + "step": 108990 + }, + { + "epoch": 31.713703811463485, + "grad_norm": 0.23301897943019867, + "learning_rate": 6.840483735509611e-07, + "loss": 0.0111, + "step": 109000 + }, + { + "epoch": 31.716613325574627, + "grad_norm": 0.12896676361560822, + "learning_rate": 6.82814845371627e-07, + "loss": 0.0082, + "step": 109010 + }, + { + "epoch": 31.719522839685773, + "grad_norm": 0.12841925024986267, + "learning_rate": 6.815824044880214e-07, + "loss": 0.0092, + "step": 109020 + }, + { + "epoch": 31.722432353796915, + "grad_norm": 0.13432614505290985, + "learning_rate": 6.803510509937394e-07, + "loss": 0.0117, + "step": 109030 + }, + { + "epoch": 31.72534186790806, + "grad_norm": 0.11475224792957306, + "learning_rate": 6.791207849822945e-07, + "loss": 0.0106, + "step": 109040 + }, + { + "epoch": 31.728251382019202, + "grad_norm": 0.13137900829315186, + "learning_rate": 6.778916065471103e-07, + "loss": 0.0131, + "step": 109050 + }, + { + "epoch": 31.731160896130348, + "grad_norm": 0.12488023936748505, + "learning_rate": 6.76663515781547e-07, + "loss": 0.0091, + "step": 109060 + }, + { + "epoch": 31.73407041024149, + "grad_norm": 0.08162689208984375, + "learning_rate": 6.754365127788636e-07, + "loss": 0.012, + "step": 109070 + }, + { + "epoch": 31.73697992435263, + "grad_norm": 0.13151663541793823, + "learning_rate": 6.742105976322405e-07, + "loss": 0.0096, + "step": 109080 + }, + { + "epoch": 31.739889438463777, + "grad_norm": 0.12818846106529236, + "learning_rate": 6.729857704347797e-07, + "loss": 0.0114, + "step": 109090 + }, + { + "epoch": 31.74279895257492, + "grad_norm": 0.09523124992847443, + "learning_rate": 6.717620312795036e-07, + "loss": 0.0099, + "step": 109100 + }, + { + "epoch": 31.745708466686064, + "grad_norm": 0.11550548672676086, + "learning_rate": 6.705393802593396e-07, + "loss": 0.01, + "step": 109110 + }, + { + "epoch": 31.748617980797206, + "grad_norm": 0.1213279515504837, + "learning_rate": 6.693178174671466e-07, + "loss": 0.0104, + "step": 109120 + }, + { + "epoch": 31.751527494908352, + "grad_norm": 0.07776990532875061, + "learning_rate": 6.680973429956888e-07, + "loss": 0.0099, + "step": 109130 + }, + { + "epoch": 31.754437009019494, + "grad_norm": 0.0968211218714714, + "learning_rate": 6.668779569376571e-07, + "loss": 0.0106, + "step": 109140 + }, + { + "epoch": 31.757346523130636, + "grad_norm": 0.08437841385602951, + "learning_rate": 6.656596593856523e-07, + "loss": 0.0093, + "step": 109150 + }, + { + "epoch": 31.76025603724178, + "grad_norm": 0.09584114700555801, + "learning_rate": 6.644424504321972e-07, + "loss": 0.0109, + "step": 109160 + }, + { + "epoch": 31.763165551352923, + "grad_norm": 0.10183250159025192, + "learning_rate": 6.632263301697345e-07, + "loss": 0.0108, + "step": 109170 + }, + { + "epoch": 31.76607506546407, + "grad_norm": 0.13275212049484253, + "learning_rate": 6.620112986906152e-07, + "loss": 0.0122, + "step": 109180 + }, + { + "epoch": 31.76898457957521, + "grad_norm": 0.17219239473342896, + "learning_rate": 6.607973560871173e-07, + "loss": 0.0099, + "step": 109190 + }, + { + "epoch": 31.771894093686356, + "grad_norm": 0.08981399238109589, + "learning_rate": 6.595845024514285e-07, + "loss": 0.0116, + "step": 109200 + }, + { + "epoch": 31.774803607797498, + "grad_norm": 0.07844484597444534, + "learning_rate": 6.583727378756571e-07, + "loss": 0.0102, + "step": 109210 + }, + { + "epoch": 31.77771312190864, + "grad_norm": 0.08904692530632019, + "learning_rate": 6.571620624518326e-07, + "loss": 0.0128, + "step": 109220 + }, + { + "epoch": 31.780622636019785, + "grad_norm": 0.17677344381809235, + "learning_rate": 6.559524762718933e-07, + "loss": 0.0106, + "step": 109230 + }, + { + "epoch": 31.783532150130927, + "grad_norm": 0.13520687818527222, + "learning_rate": 6.547439794277005e-07, + "loss": 0.0095, + "step": 109240 + }, + { + "epoch": 31.786441664242073, + "grad_norm": 0.06889946758747101, + "learning_rate": 6.535365720110326e-07, + "loss": 0.009, + "step": 109250 + }, + { + "epoch": 31.789351178353215, + "grad_norm": 0.07992031425237656, + "learning_rate": 6.523302541135845e-07, + "loss": 0.0093, + "step": 109260 + }, + { + "epoch": 31.79226069246436, + "grad_norm": 0.0660422295331955, + "learning_rate": 6.511250258269647e-07, + "loss": 0.0079, + "step": 109270 + }, + { + "epoch": 31.795170206575502, + "grad_norm": 0.1658865213394165, + "learning_rate": 6.499208872427065e-07, + "loss": 0.01, + "step": 109280 + }, + { + "epoch": 31.798079720686644, + "grad_norm": 0.07787895202636719, + "learning_rate": 6.487178384522552e-07, + "loss": 0.0088, + "step": 109290 + }, + { + "epoch": 31.80098923479779, + "grad_norm": 0.131865993142128, + "learning_rate": 6.475158795469743e-07, + "loss": 0.0091, + "step": 109300 + }, + { + "epoch": 31.80389874890893, + "grad_norm": 0.1301511824131012, + "learning_rate": 6.463150106181426e-07, + "loss": 0.0102, + "step": 109310 + }, + { + "epoch": 31.806808263020077, + "grad_norm": 0.18196681141853333, + "learning_rate": 6.451152317569587e-07, + "loss": 0.0121, + "step": 109320 + }, + { + "epoch": 31.80971777713122, + "grad_norm": 0.08443278819322586, + "learning_rate": 6.439165430545413e-07, + "loss": 0.011, + "step": 109330 + }, + { + "epoch": 31.81262729124236, + "grad_norm": 0.09034670889377594, + "learning_rate": 6.42718944601921e-07, + "loss": 0.0099, + "step": 109340 + }, + { + "epoch": 31.815536805353506, + "grad_norm": 0.10562459379434586, + "learning_rate": 6.415224364900435e-07, + "loss": 0.0112, + "step": 109350 + }, + { + "epoch": 31.81844631946465, + "grad_norm": 0.15094095468521118, + "learning_rate": 6.403270188097827e-07, + "loss": 0.0114, + "step": 109360 + }, + { + "epoch": 31.821355833575794, + "grad_norm": 0.14621177315711975, + "learning_rate": 6.391326916519192e-07, + "loss": 0.0109, + "step": 109370 + }, + { + "epoch": 31.824265347686936, + "grad_norm": 0.09624097496271133, + "learning_rate": 6.379394551071521e-07, + "loss": 0.009, + "step": 109380 + }, + { + "epoch": 31.82717486179808, + "grad_norm": 0.07902518659830093, + "learning_rate": 6.367473092661025e-07, + "loss": 0.0117, + "step": 109390 + }, + { + "epoch": 31.830084375909223, + "grad_norm": 0.12327911704778671, + "learning_rate": 6.355562542193078e-07, + "loss": 0.0094, + "step": 109400 + }, + { + "epoch": 31.832993890020365, + "grad_norm": 0.08453162759542465, + "learning_rate": 6.343662900572172e-07, + "loss": 0.0108, + "step": 109410 + }, + { + "epoch": 31.83590340413151, + "grad_norm": 0.08094175159931183, + "learning_rate": 6.33177416870202e-07, + "loss": 0.0087, + "step": 109420 + }, + { + "epoch": 31.838812918242652, + "grad_norm": 0.15256205201148987, + "learning_rate": 6.319896347485466e-07, + "loss": 0.012, + "step": 109430 + }, + { + "epoch": 31.841722432353798, + "grad_norm": 0.11254730820655823, + "learning_rate": 6.30802943782462e-07, + "loss": 0.01, + "step": 109440 + }, + { + "epoch": 31.84463194646494, + "grad_norm": 0.09835207462310791, + "learning_rate": 6.296173440620628e-07, + "loss": 0.0095, + "step": 109450 + }, + { + "epoch": 31.847541460576085, + "grad_norm": 0.13734042644500732, + "learning_rate": 6.284328356773889e-07, + "loss": 0.0092, + "step": 109460 + }, + { + "epoch": 31.850450974687227, + "grad_norm": 0.08934329450130463, + "learning_rate": 6.272494187183997e-07, + "loss": 0.0087, + "step": 109470 + }, + { + "epoch": 31.85336048879837, + "grad_norm": 0.10653985291719437, + "learning_rate": 6.260670932749635e-07, + "loss": 0.0102, + "step": 109480 + }, + { + "epoch": 31.856270002909515, + "grad_norm": 0.15239480137825012, + "learning_rate": 6.24885859436875e-07, + "loss": 0.0104, + "step": 109490 + }, + { + "epoch": 31.859179517020657, + "grad_norm": 0.09840337187051773, + "learning_rate": 6.237057172938343e-07, + "loss": 0.0096, + "step": 109500 + }, + { + "epoch": 31.862089031131802, + "grad_norm": 0.15969878435134888, + "learning_rate": 6.225266669354713e-07, + "loss": 0.0099, + "step": 109510 + }, + { + "epoch": 31.864998545242944, + "grad_norm": 0.09532878547906876, + "learning_rate": 6.213487084513242e-07, + "loss": 0.0113, + "step": 109520 + }, + { + "epoch": 31.86790805935409, + "grad_norm": 0.14271077513694763, + "learning_rate": 6.201718419308517e-07, + "loss": 0.0086, + "step": 109530 + }, + { + "epoch": 31.87081757346523, + "grad_norm": 0.14006371796131134, + "learning_rate": 6.189960674634287e-07, + "loss": 0.0107, + "step": 109540 + }, + { + "epoch": 31.873727087576373, + "grad_norm": 0.11789991706609726, + "learning_rate": 6.178213851383474e-07, + "loss": 0.01, + "step": 109550 + }, + { + "epoch": 31.87663660168752, + "grad_norm": 0.1037549152970314, + "learning_rate": 6.166477950448213e-07, + "loss": 0.0083, + "step": 109560 + }, + { + "epoch": 31.87954611579866, + "grad_norm": 0.1076866015791893, + "learning_rate": 6.154752972719691e-07, + "loss": 0.0105, + "step": 109570 + }, + { + "epoch": 31.882455629909806, + "grad_norm": 0.14943504333496094, + "learning_rate": 6.143038919088395e-07, + "loss": 0.0095, + "step": 109580 + }, + { + "epoch": 31.88536514402095, + "grad_norm": 0.09928181022405624, + "learning_rate": 6.131335790443932e-07, + "loss": 0.0095, + "step": 109590 + }, + { + "epoch": 31.88827465813209, + "grad_norm": 0.14133302867412567, + "learning_rate": 6.119643587675073e-07, + "loss": 0.011, + "step": 109600 + }, + { + "epoch": 31.891184172243236, + "grad_norm": 0.07721342146396637, + "learning_rate": 6.10796231166974e-07, + "loss": 0.0095, + "step": 109610 + }, + { + "epoch": 31.894093686354378, + "grad_norm": 0.1597939282655716, + "learning_rate": 6.096291963315059e-07, + "loss": 0.0114, + "step": 109620 + }, + { + "epoch": 31.897003200465523, + "grad_norm": 0.12081795185804367, + "learning_rate": 6.084632543497337e-07, + "loss": 0.01, + "step": 109630 + }, + { + "epoch": 31.899912714576665, + "grad_norm": 0.09710420668125153, + "learning_rate": 6.072984053102032e-07, + "loss": 0.0101, + "step": 109640 + }, + { + "epoch": 31.90282222868781, + "grad_norm": 0.07039511948823929, + "learning_rate": 6.061346493013686e-07, + "loss": 0.0102, + "step": 109650 + }, + { + "epoch": 31.905731742798952, + "grad_norm": 0.12312440574169159, + "learning_rate": 6.049719864116227e-07, + "loss": 0.0131, + "step": 109660 + }, + { + "epoch": 31.908641256910094, + "grad_norm": 0.08751139044761658, + "learning_rate": 6.038104167292546e-07, + "loss": 0.0105, + "step": 109670 + }, + { + "epoch": 31.91155077102124, + "grad_norm": 0.125532865524292, + "learning_rate": 6.026499403424757e-07, + "loss": 0.0118, + "step": 109680 + }, + { + "epoch": 31.91446028513238, + "grad_norm": 0.09237536042928696, + "learning_rate": 6.014905573394202e-07, + "loss": 0.0107, + "step": 109690 + }, + { + "epoch": 31.917369799243527, + "grad_norm": 0.08037266880273819, + "learning_rate": 6.00332267808138e-07, + "loss": 0.0107, + "step": 109700 + }, + { + "epoch": 31.92027931335467, + "grad_norm": 0.13146348297595978, + "learning_rate": 5.991750718365885e-07, + "loss": 0.0099, + "step": 109710 + }, + { + "epoch": 31.923188827465815, + "grad_norm": 0.10408502072095871, + "learning_rate": 5.980189695126547e-07, + "loss": 0.0093, + "step": 109720 + }, + { + "epoch": 31.926098341576957, + "grad_norm": 0.10190797597169876, + "learning_rate": 5.968639609241349e-07, + "loss": 0.0105, + "step": 109730 + }, + { + "epoch": 31.9290078556881, + "grad_norm": 0.12258736789226532, + "learning_rate": 5.957100461587472e-07, + "loss": 0.0107, + "step": 109740 + }, + { + "epoch": 31.931917369799244, + "grad_norm": 0.09874217212200165, + "learning_rate": 5.945572253041198e-07, + "loss": 0.0089, + "step": 109750 + }, + { + "epoch": 31.934826883910386, + "grad_norm": 0.1533600389957428, + "learning_rate": 5.934054984478026e-07, + "loss": 0.0097, + "step": 109760 + }, + { + "epoch": 31.93773639802153, + "grad_norm": 0.09090138226747513, + "learning_rate": 5.922548656772641e-07, + "loss": 0.01, + "step": 109770 + }, + { + "epoch": 31.940645912132673, + "grad_norm": 0.11029686778783798, + "learning_rate": 5.911053270798861e-07, + "loss": 0.0106, + "step": 109780 + }, + { + "epoch": 31.94355542624382, + "grad_norm": 0.1376999318599701, + "learning_rate": 5.899568827429668e-07, + "loss": 0.0091, + "step": 109790 + }, + { + "epoch": 31.94646494035496, + "grad_norm": 0.11007507145404816, + "learning_rate": 5.888095327537235e-07, + "loss": 0.0094, + "step": 109800 + }, + { + "epoch": 31.949374454466103, + "grad_norm": 0.10399533808231354, + "learning_rate": 5.876632771992929e-07, + "loss": 0.0095, + "step": 109810 + }, + { + "epoch": 31.952283968577248, + "grad_norm": 0.0990951731801033, + "learning_rate": 5.865181161667221e-07, + "loss": 0.0092, + "step": 109820 + }, + { + "epoch": 31.95519348268839, + "grad_norm": 0.1413932591676712, + "learning_rate": 5.8537404974298e-07, + "loss": 0.0101, + "step": 109830 + }, + { + "epoch": 31.958102996799536, + "grad_norm": 0.11528007686138153, + "learning_rate": 5.842310780149502e-07, + "loss": 0.0111, + "step": 109840 + }, + { + "epoch": 31.961012510910678, + "grad_norm": 0.07328586280345917, + "learning_rate": 5.830892010694333e-07, + "loss": 0.0124, + "step": 109850 + }, + { + "epoch": 31.963922025021823, + "grad_norm": 0.096938356757164, + "learning_rate": 5.819484189931501e-07, + "loss": 0.0097, + "step": 109860 + }, + { + "epoch": 31.966831539132965, + "grad_norm": 0.13627147674560547, + "learning_rate": 5.808087318727329e-07, + "loss": 0.0102, + "step": 109870 + }, + { + "epoch": 31.969741053244107, + "grad_norm": 0.10983692854642868, + "learning_rate": 5.796701397947341e-07, + "loss": 0.0106, + "step": 109880 + }, + { + "epoch": 31.972650567355252, + "grad_norm": 0.06913948804140091, + "learning_rate": 5.785326428456211e-07, + "loss": 0.0086, + "step": 109890 + }, + { + "epoch": 31.975560081466394, + "grad_norm": 0.10427329689264297, + "learning_rate": 5.773962411117834e-07, + "loss": 0.0109, + "step": 109900 + }, + { + "epoch": 31.97846959557754, + "grad_norm": 0.09712781757116318, + "learning_rate": 5.762609346795184e-07, + "loss": 0.0143, + "step": 109910 + }, + { + "epoch": 31.98137910968868, + "grad_norm": 0.09159762412309647, + "learning_rate": 5.751267236350454e-07, + "loss": 0.0108, + "step": 109920 + }, + { + "epoch": 31.984288623799827, + "grad_norm": 0.10268209874629974, + "learning_rate": 5.739936080645041e-07, + "loss": 0.0103, + "step": 109930 + }, + { + "epoch": 31.98719813791097, + "grad_norm": 0.0755164697766304, + "learning_rate": 5.728615880539456e-07, + "loss": 0.0103, + "step": 109940 + }, + { + "epoch": 31.99010765202211, + "grad_norm": 0.0887598842382431, + "learning_rate": 5.717306636893344e-07, + "loss": 0.0104, + "step": 109950 + }, + { + "epoch": 31.993017166133257, + "grad_norm": 0.0940188392996788, + "learning_rate": 5.70600835056565e-07, + "loss": 0.0096, + "step": 109960 + }, + { + "epoch": 31.9959266802444, + "grad_norm": 0.3889932632446289, + "learning_rate": 5.694721022414373e-07, + "loss": 0.0113, + "step": 109970 + }, + { + "epoch": 31.998836194355544, + "grad_norm": 0.07307486236095428, + "learning_rate": 5.683444653296676e-07, + "loss": 0.0075, + "step": 109980 + }, + { + "epoch": 32.001745708466686, + "grad_norm": 0.1480894535779953, + "learning_rate": 5.672179244068959e-07, + "loss": 0.0089, + "step": 109990 + }, + { + "epoch": 32.00465522257783, + "grad_norm": 0.09260493516921997, + "learning_rate": 5.660924795586753e-07, + "loss": 0.0126, + "step": 110000 + }, + { + "epoch": 32.00756473668897, + "grad_norm": 0.10900694876909256, + "learning_rate": 5.649681308704758e-07, + "loss": 0.0095, + "step": 110010 + }, + { + "epoch": 32.010474250800115, + "grad_norm": 0.3491433262825012, + "learning_rate": 5.638448784276822e-07, + "loss": 0.01, + "step": 110020 + }, + { + "epoch": 32.01338376491126, + "grad_norm": 0.12995365262031555, + "learning_rate": 5.627227223156001e-07, + "loss": 0.0116, + "step": 110030 + }, + { + "epoch": 32.016293279022406, + "grad_norm": 0.0968073382973671, + "learning_rate": 5.616016626194492e-07, + "loss": 0.0103, + "step": 110040 + }, + { + "epoch": 32.019202793133545, + "grad_norm": 0.11376890540122986, + "learning_rate": 5.604816994243683e-07, + "loss": 0.0093, + "step": 110050 + }, + { + "epoch": 32.02211230724469, + "grad_norm": 0.08295627683401108, + "learning_rate": 5.593628328154077e-07, + "loss": 0.0088, + "step": 110060 + }, + { + "epoch": 32.025021821355836, + "grad_norm": 0.09366970509290695, + "learning_rate": 5.582450628775426e-07, + "loss": 0.0107, + "step": 110070 + }, + { + "epoch": 32.027931335466974, + "grad_norm": 0.16708675026893616, + "learning_rate": 5.571283896956569e-07, + "loss": 0.0107, + "step": 110080 + }, + { + "epoch": 32.03084084957812, + "grad_norm": 0.09746364504098892, + "learning_rate": 5.560128133545545e-07, + "loss": 0.0096, + "step": 110090 + }, + { + "epoch": 32.033750363689265, + "grad_norm": 0.09079255908727646, + "learning_rate": 5.548983339389557e-07, + "loss": 0.0101, + "step": 110100 + }, + { + "epoch": 32.03665987780041, + "grad_norm": 0.11083856970071793, + "learning_rate": 5.537849515335014e-07, + "loss": 0.0107, + "step": 110110 + }, + { + "epoch": 32.03956939191155, + "grad_norm": 0.07990050315856934, + "learning_rate": 5.526726662227421e-07, + "loss": 0.0104, + "step": 110120 + }, + { + "epoch": 32.042478906022694, + "grad_norm": 0.09091328084468842, + "learning_rate": 5.515614780911521e-07, + "loss": 0.0104, + "step": 110130 + }, + { + "epoch": 32.04538842013384, + "grad_norm": 0.09374379366636276, + "learning_rate": 5.504513872231137e-07, + "loss": 0.0113, + "step": 110140 + }, + { + "epoch": 32.04829793424498, + "grad_norm": 0.15270808339118958, + "learning_rate": 5.493423937029346e-07, + "loss": 0.0096, + "step": 110150 + }, + { + "epoch": 32.051207448356124, + "grad_norm": 0.13539308309555054, + "learning_rate": 5.48234497614834e-07, + "loss": 0.012, + "step": 110160 + }, + { + "epoch": 32.05411696246727, + "grad_norm": 0.10375829041004181, + "learning_rate": 5.471276990429497e-07, + "loss": 0.01, + "step": 110170 + }, + { + "epoch": 32.057026476578415, + "grad_norm": 0.1230444461107254, + "learning_rate": 5.460219980713377e-07, + "loss": 0.011, + "step": 110180 + }, + { + "epoch": 32.05993599068955, + "grad_norm": 0.11785105615854263, + "learning_rate": 5.449173947839659e-07, + "loss": 0.0099, + "step": 110190 + }, + { + "epoch": 32.0628455048007, + "grad_norm": 0.10912847518920898, + "learning_rate": 5.438138892647237e-07, + "loss": 0.0108, + "step": 110200 + }, + { + "epoch": 32.065755018911844, + "grad_norm": 0.1549524962902069, + "learning_rate": 5.427114815974127e-07, + "loss": 0.0097, + "step": 110210 + }, + { + "epoch": 32.06866453302298, + "grad_norm": 0.14858147501945496, + "learning_rate": 5.416101718657556e-07, + "loss": 0.0104, + "step": 110220 + }, + { + "epoch": 32.07157404713413, + "grad_norm": 0.07729260623455048, + "learning_rate": 5.405099601533892e-07, + "loss": 0.0117, + "step": 110230 + }, + { + "epoch": 32.07448356124527, + "grad_norm": 0.11935745924711227, + "learning_rate": 5.394108465438679e-07, + "loss": 0.0102, + "step": 110240 + }, + { + "epoch": 32.07739307535642, + "grad_norm": 0.10141675174236298, + "learning_rate": 5.383128311206586e-07, + "loss": 0.0109, + "step": 110250 + }, + { + "epoch": 32.08030258946756, + "grad_norm": 0.1124369353055954, + "learning_rate": 5.372159139671512e-07, + "loss": 0.0103, + "step": 110260 + }, + { + "epoch": 32.0832121035787, + "grad_norm": 0.1011970043182373, + "learning_rate": 5.361200951666522e-07, + "loss": 0.0087, + "step": 110270 + }, + { + "epoch": 32.08612161768985, + "grad_norm": 0.08529413491487503, + "learning_rate": 5.350253748023754e-07, + "loss": 0.0105, + "step": 110280 + }, + { + "epoch": 32.089031131800986, + "grad_norm": 0.09971300512552261, + "learning_rate": 5.339317529574606e-07, + "loss": 0.0093, + "step": 110290 + }, + { + "epoch": 32.09194064591213, + "grad_norm": 0.07479105144739151, + "learning_rate": 5.328392297149647e-07, + "loss": 0.0099, + "step": 110300 + }, + { + "epoch": 32.09485016002328, + "grad_norm": 0.11990593373775482, + "learning_rate": 5.317478051578534e-07, + "loss": 0.0105, + "step": 110310 + }, + { + "epoch": 32.09775967413442, + "grad_norm": 0.08928803354501724, + "learning_rate": 5.306574793690117e-07, + "loss": 0.0111, + "step": 110320 + }, + { + "epoch": 32.10066918824556, + "grad_norm": 0.08716239780187607, + "learning_rate": 5.295682524312467e-07, + "loss": 0.0105, + "step": 110330 + }, + { + "epoch": 32.10357870235671, + "grad_norm": 0.07820849120616913, + "learning_rate": 5.284801244272774e-07, + "loss": 0.0099, + "step": 110340 + }, + { + "epoch": 32.10648821646785, + "grad_norm": 0.08609037101268768, + "learning_rate": 5.273930954397393e-07, + "loss": 0.01, + "step": 110350 + }, + { + "epoch": 32.10939773057899, + "grad_norm": 0.08315495401620865, + "learning_rate": 5.263071655511798e-07, + "loss": 0.0099, + "step": 110360 + }, + { + "epoch": 32.112307244690136, + "grad_norm": 0.11264033615589142, + "learning_rate": 5.252223348440776e-07, + "loss": 0.0105, + "step": 110370 + }, + { + "epoch": 32.11521675880128, + "grad_norm": 0.07476692646741867, + "learning_rate": 5.241386034008139e-07, + "loss": 0.0085, + "step": 110380 + }, + { + "epoch": 32.11812627291243, + "grad_norm": 0.11191030591726303, + "learning_rate": 5.230559713036892e-07, + "loss": 0.0103, + "step": 110390 + }, + { + "epoch": 32.121035787023565, + "grad_norm": 0.1089220717549324, + "learning_rate": 5.219744386349229e-07, + "loss": 0.0123, + "step": 110400 + }, + { + "epoch": 32.12394530113471, + "grad_norm": 0.1088428869843483, + "learning_rate": 5.208940054766543e-07, + "loss": 0.0084, + "step": 110410 + }, + { + "epoch": 32.126854815245856, + "grad_norm": 0.07637573778629303, + "learning_rate": 5.198146719109309e-07, + "loss": 0.0111, + "step": 110420 + }, + { + "epoch": 32.129764329356995, + "grad_norm": 0.07246862351894379, + "learning_rate": 5.187364380197224e-07, + "loss": 0.0113, + "step": 110430 + }, + { + "epoch": 32.13267384346814, + "grad_norm": 0.11853985488414764, + "learning_rate": 5.176593038849131e-07, + "loss": 0.0108, + "step": 110440 + }, + { + "epoch": 32.135583357579286, + "grad_norm": 0.1336696743965149, + "learning_rate": 5.165832695883043e-07, + "loss": 0.0129, + "step": 110450 + }, + { + "epoch": 32.13849287169043, + "grad_norm": 0.10671409219503403, + "learning_rate": 5.155083352116141e-07, + "loss": 0.0102, + "step": 110460 + }, + { + "epoch": 32.14140238580157, + "grad_norm": 0.11619283258914948, + "learning_rate": 5.144345008364753e-07, + "loss": 0.0112, + "step": 110470 + }, + { + "epoch": 32.144311899912715, + "grad_norm": 0.10700175911188126, + "learning_rate": 5.133617665444429e-07, + "loss": 0.0107, + "step": 110480 + }, + { + "epoch": 32.14722141402386, + "grad_norm": 0.10344427078962326, + "learning_rate": 5.122901324169782e-07, + "loss": 0.0098, + "step": 110490 + }, + { + "epoch": 32.150130928135, + "grad_norm": 0.11591722071170807, + "learning_rate": 5.112195985354695e-07, + "loss": 0.0101, + "step": 110500 + }, + { + "epoch": 32.153040442246144, + "grad_norm": 0.05743898078799248, + "learning_rate": 5.101501649812118e-07, + "loss": 0.009, + "step": 110510 + }, + { + "epoch": 32.15594995635729, + "grad_norm": 0.08118270337581635, + "learning_rate": 5.090818318354285e-07, + "loss": 0.0103, + "step": 110520 + }, + { + "epoch": 32.15885947046843, + "grad_norm": 0.12732477486133575, + "learning_rate": 5.080145991792445e-07, + "loss": 0.0132, + "step": 110530 + }, + { + "epoch": 32.161768984579574, + "grad_norm": 0.08107178658246994, + "learning_rate": 5.069484670937152e-07, + "loss": 0.0105, + "step": 110540 + }, + { + "epoch": 32.16467849869072, + "grad_norm": 0.10749290138483047, + "learning_rate": 5.058834356598024e-07, + "loss": 0.0107, + "step": 110550 + }, + { + "epoch": 32.167588012801865, + "grad_norm": 0.09776169061660767, + "learning_rate": 5.048195049583898e-07, + "loss": 0.01, + "step": 110560 + }, + { + "epoch": 32.170497526913, + "grad_norm": 0.10464395582675934, + "learning_rate": 5.037566750702777e-07, + "loss": 0.0104, + "step": 110570 + }, + { + "epoch": 32.17340704102415, + "grad_norm": 0.10297301411628723, + "learning_rate": 5.026949460761765e-07, + "loss": 0.0098, + "step": 110580 + }, + { + "epoch": 32.176316555135294, + "grad_norm": 0.1304662525653839, + "learning_rate": 5.016343180567217e-07, + "loss": 0.0109, + "step": 110590 + }, + { + "epoch": 32.17922606924643, + "grad_norm": 0.07984668016433716, + "learning_rate": 5.005747910924607e-07, + "loss": 0.0087, + "step": 110600 + }, + { + "epoch": 32.18213558335758, + "grad_norm": 0.07351759076118469, + "learning_rate": 4.995163652638556e-07, + "loss": 0.0089, + "step": 110610 + }, + { + "epoch": 32.18504509746872, + "grad_norm": 0.06648815423250198, + "learning_rate": 4.984590406512874e-07, + "loss": 0.0089, + "step": 110620 + }, + { + "epoch": 32.18795461157987, + "grad_norm": 0.1430829018354416, + "learning_rate": 4.974028173350531e-07, + "loss": 0.0117, + "step": 110630 + }, + { + "epoch": 32.19086412569101, + "grad_norm": 0.11194227635860443, + "learning_rate": 4.963476953953672e-07, + "loss": 0.009, + "step": 110640 + }, + { + "epoch": 32.19377363980215, + "grad_norm": 0.10054033249616623, + "learning_rate": 4.952936749123572e-07, + "loss": 0.0102, + "step": 110650 + }, + { + "epoch": 32.1966831539133, + "grad_norm": 0.10551789402961731, + "learning_rate": 4.942407559660689e-07, + "loss": 0.0106, + "step": 110660 + }, + { + "epoch": 32.19959266802444, + "grad_norm": 0.08278116583824158, + "learning_rate": 4.931889386364669e-07, + "loss": 0.0097, + "step": 110670 + }, + { + "epoch": 32.20250218213558, + "grad_norm": 0.12361503392457962, + "learning_rate": 4.921382230034288e-07, + "loss": 0.0111, + "step": 110680 + }, + { + "epoch": 32.20541169624673, + "grad_norm": 0.106845922768116, + "learning_rate": 4.910886091467476e-07, + "loss": 0.0113, + "step": 110690 + }, + { + "epoch": 32.20832121035787, + "grad_norm": 0.19084632396697998, + "learning_rate": 4.900400971461361e-07, + "loss": 0.0112, + "step": 110700 + }, + { + "epoch": 32.21123072446901, + "grad_norm": 0.13887165486812592, + "learning_rate": 4.889926870812222e-07, + "loss": 0.0104, + "step": 110710 + }, + { + "epoch": 32.21414023858016, + "grad_norm": 0.08551866561174393, + "learning_rate": 4.879463790315508e-07, + "loss": 0.0114, + "step": 110720 + }, + { + "epoch": 32.2170497526913, + "grad_norm": 0.07743681967258453, + "learning_rate": 4.869011730765799e-07, + "loss": 0.0097, + "step": 110730 + }, + { + "epoch": 32.21995926680244, + "grad_norm": 0.09723922610282898, + "learning_rate": 4.858570692956843e-07, + "loss": 0.0093, + "step": 110740 + }, + { + "epoch": 32.222868780913586, + "grad_norm": 0.12477528303861618, + "learning_rate": 4.848140677681623e-07, + "loss": 0.0103, + "step": 110750 + }, + { + "epoch": 32.22577829502473, + "grad_norm": 0.09990587830543518, + "learning_rate": 4.837721685732188e-07, + "loss": 0.0093, + "step": 110760 + }, + { + "epoch": 32.22868780913588, + "grad_norm": 0.06492569297552109, + "learning_rate": 4.827313717899807e-07, + "loss": 0.0117, + "step": 110770 + }, + { + "epoch": 32.231597323247016, + "grad_norm": 0.11801129579544067, + "learning_rate": 4.816916774974894e-07, + "loss": 0.0098, + "step": 110780 + }, + { + "epoch": 32.23450683735816, + "grad_norm": 0.09648147225379944, + "learning_rate": 4.806530857747021e-07, + "loss": 0.0128, + "step": 110790 + }, + { + "epoch": 32.23741635146931, + "grad_norm": 0.10513093322515488, + "learning_rate": 4.796155967004939e-07, + "loss": 0.012, + "step": 110800 + }, + { + "epoch": 32.240325865580445, + "grad_norm": 0.09351974725723267, + "learning_rate": 4.785792103536551e-07, + "loss": 0.012, + "step": 110810 + }, + { + "epoch": 32.24323537969159, + "grad_norm": 0.10203740000724792, + "learning_rate": 4.775439268128928e-07, + "loss": 0.0109, + "step": 110820 + }, + { + "epoch": 32.246144893802736, + "grad_norm": 0.08104202151298523, + "learning_rate": 4.765097461568274e-07, + "loss": 0.0114, + "step": 110830 + }, + { + "epoch": 32.24905440791388, + "grad_norm": 0.09138374775648117, + "learning_rate": 4.7547666846400275e-07, + "loss": 0.0115, + "step": 110840 + }, + { + "epoch": 32.25196392202502, + "grad_norm": 0.09559999406337738, + "learning_rate": 4.744446938128694e-07, + "loss": 0.011, + "step": 110850 + }, + { + "epoch": 32.254873436136165, + "grad_norm": 0.12647870182991028, + "learning_rate": 4.734138222818013e-07, + "loss": 0.0089, + "step": 110860 + }, + { + "epoch": 32.25778295024731, + "grad_norm": 0.08500046283006668, + "learning_rate": 4.723840539490876e-07, + "loss": 0.012, + "step": 110870 + }, + { + "epoch": 32.26069246435845, + "grad_norm": 0.11937836557626724, + "learning_rate": 4.713553888929306e-07, + "loss": 0.0099, + "step": 110880 + }, + { + "epoch": 32.263601978469595, + "grad_norm": 0.11506782472133636, + "learning_rate": 4.703278271914496e-07, + "loss": 0.0112, + "step": 110890 + }, + { + "epoch": 32.26651149258074, + "grad_norm": 0.08518647402524948, + "learning_rate": 4.6930136892268547e-07, + "loss": 0.0104, + "step": 110900 + }, + { + "epoch": 32.269421006691886, + "grad_norm": 0.19467616081237793, + "learning_rate": 4.6827601416458756e-07, + "loss": 0.0105, + "step": 110910 + }, + { + "epoch": 32.272330520803024, + "grad_norm": 0.07783202081918716, + "learning_rate": 4.6725176299502346e-07, + "loss": 0.008, + "step": 110920 + }, + { + "epoch": 32.27524003491417, + "grad_norm": 0.1379561573266983, + "learning_rate": 4.662286154917794e-07, + "loss": 0.0092, + "step": 110930 + }, + { + "epoch": 32.278149549025315, + "grad_norm": 0.11104656010866165, + "learning_rate": 4.6520657173256155e-07, + "loss": 0.0111, + "step": 110940 + }, + { + "epoch": 32.28105906313645, + "grad_norm": 0.0926966443657875, + "learning_rate": 4.641856317949811e-07, + "loss": 0.0088, + "step": 110950 + }, + { + "epoch": 32.2839685772476, + "grad_norm": 0.08550000935792923, + "learning_rate": 4.63165795756571e-07, + "loss": 0.012, + "step": 110960 + }, + { + "epoch": 32.286878091358744, + "grad_norm": 0.12177462875843048, + "learning_rate": 4.6214706369478775e-07, + "loss": 0.0089, + "step": 110970 + }, + { + "epoch": 32.28978760546989, + "grad_norm": 0.10299546271562576, + "learning_rate": 4.611294356869944e-07, + "loss": 0.01, + "step": 110980 + }, + { + "epoch": 32.29269711958103, + "grad_norm": 0.12372317910194397, + "learning_rate": 4.601129118104691e-07, + "loss": 0.0101, + "step": 110990 + }, + { + "epoch": 32.295606633692174, + "grad_norm": 0.09771998971700668, + "learning_rate": 4.590974921424135e-07, + "loss": 0.0105, + "step": 111000 + }, + { + "epoch": 32.29851614780332, + "grad_norm": 0.16848602890968323, + "learning_rate": 4.5808317675994416e-07, + "loss": 0.0103, + "step": 111010 + }, + { + "epoch": 32.30142566191446, + "grad_norm": 0.07659286260604858, + "learning_rate": 4.5706996574008787e-07, + "loss": 0.009, + "step": 111020 + }, + { + "epoch": 32.3043351760256, + "grad_norm": 0.11169321089982986, + "learning_rate": 4.560578591597914e-07, + "loss": 0.0128, + "step": 111030 + }, + { + "epoch": 32.30724469013675, + "grad_norm": 0.1049814373254776, + "learning_rate": 4.550468570959199e-07, + "loss": 0.0117, + "step": 111040 + }, + { + "epoch": 32.310154204247894, + "grad_norm": 0.10528922826051712, + "learning_rate": 4.540369596252519e-07, + "loss": 0.0092, + "step": 111050 + }, + { + "epoch": 32.31306371835903, + "grad_norm": 0.10050978511571884, + "learning_rate": 4.530281668244812e-07, + "loss": 0.0111, + "step": 111060 + }, + { + "epoch": 32.31597323247018, + "grad_norm": 0.12983249127864838, + "learning_rate": 4.520204787702181e-07, + "loss": 0.0095, + "step": 111070 + }, + { + "epoch": 32.31888274658132, + "grad_norm": 0.21211929619312286, + "learning_rate": 4.5101389553899476e-07, + "loss": 0.0115, + "step": 111080 + }, + { + "epoch": 32.32179226069246, + "grad_norm": 0.11572299897670746, + "learning_rate": 4.500084172072499e-07, + "loss": 0.01, + "step": 111090 + }, + { + "epoch": 32.32470177480361, + "grad_norm": 0.07481752336025238, + "learning_rate": 4.4900404385134266e-07, + "loss": 0.009, + "step": 111100 + }, + { + "epoch": 32.32761128891475, + "grad_norm": 0.16645170748233795, + "learning_rate": 4.4800077554755015e-07, + "loss": 0.0097, + "step": 111110 + }, + { + "epoch": 32.3305208030259, + "grad_norm": 0.11020021885633469, + "learning_rate": 4.4699861237206654e-07, + "loss": 0.01, + "step": 111120 + }, + { + "epoch": 32.33343031713704, + "grad_norm": 0.17686229944229126, + "learning_rate": 4.459975544009942e-07, + "loss": 0.0126, + "step": 111130 + }, + { + "epoch": 32.33633983124818, + "grad_norm": 0.09133900701999664, + "learning_rate": 4.4499760171036075e-07, + "loss": 0.0103, + "step": 111140 + }, + { + "epoch": 32.33924934535933, + "grad_norm": 0.12554562091827393, + "learning_rate": 4.4399875437610374e-07, + "loss": 0.0108, + "step": 111150 + }, + { + "epoch": 32.342158859470466, + "grad_norm": 0.12662529945373535, + "learning_rate": 4.4300101247408074e-07, + "loss": 0.0091, + "step": 111160 + }, + { + "epoch": 32.34506837358161, + "grad_norm": 0.07981674373149872, + "learning_rate": 4.4200437608006294e-07, + "loss": 0.0102, + "step": 111170 + }, + { + "epoch": 32.34797788769276, + "grad_norm": 0.15082454681396484, + "learning_rate": 4.4100884526973806e-07, + "loss": 0.0101, + "step": 111180 + }, + { + "epoch": 32.3508874018039, + "grad_norm": 0.11536887288093567, + "learning_rate": 4.400144201187123e-07, + "loss": 0.011, + "step": 111190 + }, + { + "epoch": 32.35379691591504, + "grad_norm": 0.11873277276754379, + "learning_rate": 4.390211007025002e-07, + "loss": 0.0096, + "step": 111200 + }, + { + "epoch": 32.356706430026186, + "grad_norm": 0.12898659706115723, + "learning_rate": 4.3802888709654487e-07, + "loss": 0.0082, + "step": 111210 + }, + { + "epoch": 32.35961594413733, + "grad_norm": 0.09966445714235306, + "learning_rate": 4.3703777937619263e-07, + "loss": 0.0105, + "step": 111220 + }, + { + "epoch": 32.36252545824847, + "grad_norm": 0.12183159589767456, + "learning_rate": 4.3604777761671487e-07, + "loss": 0.0103, + "step": 111230 + }, + { + "epoch": 32.365434972359616, + "grad_norm": 0.08176948130130768, + "learning_rate": 4.350588818932949e-07, + "loss": 0.0093, + "step": 111240 + }, + { + "epoch": 32.36834448647076, + "grad_norm": 0.09401549398899078, + "learning_rate": 4.340710922810326e-07, + "loss": 0.0094, + "step": 111250 + }, + { + "epoch": 32.3712540005819, + "grad_norm": 0.09176450222730637, + "learning_rate": 4.330844088549413e-07, + "loss": 0.0111, + "step": 111260 + }, + { + "epoch": 32.374163514693045, + "grad_norm": 0.12290462851524353, + "learning_rate": 4.320988316899593e-07, + "loss": 0.0088, + "step": 111270 + }, + { + "epoch": 32.37707302880419, + "grad_norm": 0.1038379818201065, + "learning_rate": 4.311143608609319e-07, + "loss": 0.0112, + "step": 111280 + }, + { + "epoch": 32.379982542915336, + "grad_norm": 0.11723941564559937, + "learning_rate": 4.301309964426192e-07, + "loss": 0.0098, + "step": 111290 + }, + { + "epoch": 32.382892057026474, + "grad_norm": 0.15048016607761383, + "learning_rate": 4.291487385097065e-07, + "loss": 0.0108, + "step": 111300 + }, + { + "epoch": 32.38580157113762, + "grad_norm": 0.13045470416545868, + "learning_rate": 4.2816758713678903e-07, + "loss": 0.0127, + "step": 111310 + }, + { + "epoch": 32.388711085248765, + "grad_norm": 0.12080246210098267, + "learning_rate": 4.2718754239837567e-07, + "loss": 0.0105, + "step": 111320 + }, + { + "epoch": 32.391620599359904, + "grad_norm": 0.06512724608182907, + "learning_rate": 4.2620860436889674e-07, + "loss": 0.0096, + "step": 111330 + }, + { + "epoch": 32.39453011347105, + "grad_norm": 0.08803107589483261, + "learning_rate": 4.252307731226945e-07, + "loss": 0.0101, + "step": 111340 + }, + { + "epoch": 32.397439627582195, + "grad_norm": 0.11606394499540329, + "learning_rate": 4.2425404873403295e-07, + "loss": 0.0102, + "step": 111350 + }, + { + "epoch": 32.40034914169334, + "grad_norm": 0.08336324989795685, + "learning_rate": 4.232784312770827e-07, + "loss": 0.0093, + "step": 111360 + }, + { + "epoch": 32.40325865580448, + "grad_norm": 0.06973573565483093, + "learning_rate": 4.2230392082593605e-07, + "loss": 0.0091, + "step": 111370 + }, + { + "epoch": 32.406168169915624, + "grad_norm": 0.08586416393518448, + "learning_rate": 4.213305174546056e-07, + "loss": 0.0088, + "step": 111380 + }, + { + "epoch": 32.40907768402677, + "grad_norm": 0.11824632436037064, + "learning_rate": 4.203582212370105e-07, + "loss": 0.0117, + "step": 111390 + }, + { + "epoch": 32.41198719813791, + "grad_norm": 0.08724639564752579, + "learning_rate": 4.1938703224699e-07, + "loss": 0.0103, + "step": 111400 + }, + { + "epoch": 32.41489671224905, + "grad_norm": 0.10601670295000076, + "learning_rate": 4.184169505583002e-07, + "loss": 0.0143, + "step": 111410 + }, + { + "epoch": 32.4178062263602, + "grad_norm": 0.09941036254167557, + "learning_rate": 4.1744797624461537e-07, + "loss": 0.0111, + "step": 111420 + }, + { + "epoch": 32.420715740471344, + "grad_norm": 0.09278449416160583, + "learning_rate": 4.164801093795184e-07, + "loss": 0.0099, + "step": 111430 + }, + { + "epoch": 32.42362525458248, + "grad_norm": 0.09807746857404709, + "learning_rate": 4.1551335003651545e-07, + "loss": 0.0119, + "step": 111440 + }, + { + "epoch": 32.42653476869363, + "grad_norm": 0.08589839935302734, + "learning_rate": 4.145476982890228e-07, + "loss": 0.01, + "step": 111450 + }, + { + "epoch": 32.42944428280477, + "grad_norm": 0.12116020917892456, + "learning_rate": 4.135831542103785e-07, + "loss": 0.0107, + "step": 111460 + }, + { + "epoch": 32.43235379691591, + "grad_norm": 0.1271299570798874, + "learning_rate": 4.1261971787383056e-07, + "loss": 0.0105, + "step": 111470 + }, + { + "epoch": 32.43526331102706, + "grad_norm": 0.10649609565734863, + "learning_rate": 4.1165738935254716e-07, + "loss": 0.0098, + "step": 111480 + }, + { + "epoch": 32.4381728251382, + "grad_norm": 0.10874376446008682, + "learning_rate": 4.1069616871961147e-07, + "loss": 0.0117, + "step": 111490 + }, + { + "epoch": 32.44108233924935, + "grad_norm": 0.12766917049884796, + "learning_rate": 4.0973605604801846e-07, + "loss": 0.0103, + "step": 111500 + }, + { + "epoch": 32.44399185336049, + "grad_norm": 0.12993252277374268, + "learning_rate": 4.0877705141068645e-07, + "loss": 0.0099, + "step": 111510 + }, + { + "epoch": 32.44690136747163, + "grad_norm": 0.1120728850364685, + "learning_rate": 4.078191548804422e-07, + "loss": 0.0107, + "step": 111520 + }, + { + "epoch": 32.44981088158278, + "grad_norm": 0.10824897140264511, + "learning_rate": 4.068623665300342e-07, + "loss": 0.0117, + "step": 111530 + }, + { + "epoch": 32.452720395693916, + "grad_norm": 0.10666760802268982, + "learning_rate": 4.059066864321226e-07, + "loss": 0.01, + "step": 111540 + }, + { + "epoch": 32.45562990980506, + "grad_norm": 0.10978882759809494, + "learning_rate": 4.049521146592877e-07, + "loss": 0.012, + "step": 111550 + }, + { + "epoch": 32.45853942391621, + "grad_norm": 0.12022983282804489, + "learning_rate": 4.0399865128401824e-07, + "loss": 0.0113, + "step": 111560 + }, + { + "epoch": 32.46144893802735, + "grad_norm": 0.12384084612131119, + "learning_rate": 4.030462963787246e-07, + "loss": 0.0097, + "step": 111570 + }, + { + "epoch": 32.46435845213849, + "grad_norm": 0.1210019513964653, + "learning_rate": 4.020950500157372e-07, + "loss": 0.0093, + "step": 111580 + }, + { + "epoch": 32.467267966249636, + "grad_norm": 0.0940258800983429, + "learning_rate": 4.0114491226729e-07, + "loss": 0.0108, + "step": 111590 + }, + { + "epoch": 32.47017748036078, + "grad_norm": 0.11296091973781586, + "learning_rate": 4.0019588320554355e-07, + "loss": 0.0109, + "step": 111600 + }, + { + "epoch": 32.47308699447192, + "grad_norm": 0.09578882157802582, + "learning_rate": 3.9924796290257025e-07, + "loss": 0.0098, + "step": 111610 + }, + { + "epoch": 32.475996508583066, + "grad_norm": 0.084278903901577, + "learning_rate": 3.983011514303575e-07, + "loss": 0.0089, + "step": 111620 + }, + { + "epoch": 32.47890602269421, + "grad_norm": 0.10463351756334305, + "learning_rate": 3.973554488608078e-07, + "loss": 0.0092, + "step": 111630 + }, + { + "epoch": 32.48181553680536, + "grad_norm": 0.08153869956731796, + "learning_rate": 3.9641085526574364e-07, + "loss": 0.0108, + "step": 111640 + }, + { + "epoch": 32.484725050916495, + "grad_norm": 0.1398952603340149, + "learning_rate": 3.9546737071690097e-07, + "loss": 0.0106, + "step": 111650 + }, + { + "epoch": 32.48763456502764, + "grad_norm": 0.10244506597518921, + "learning_rate": 3.9452499528592924e-07, + "loss": 0.0086, + "step": 111660 + }, + { + "epoch": 32.490544079138786, + "grad_norm": 0.12245788425207138, + "learning_rate": 3.9358372904439277e-07, + "loss": 0.0102, + "step": 111670 + }, + { + "epoch": 32.493453593249924, + "grad_norm": 0.214147686958313, + "learning_rate": 3.926435720637828e-07, + "loss": 0.0113, + "step": 111680 + }, + { + "epoch": 32.49636310736107, + "grad_norm": 0.1188432052731514, + "learning_rate": 3.917045244154921e-07, + "loss": 0.01, + "step": 111690 + }, + { + "epoch": 32.499272621472215, + "grad_norm": 0.14390525221824646, + "learning_rate": 3.907665861708354e-07, + "loss": 0.0087, + "step": 111700 + }, + { + "epoch": 32.50218213558336, + "grad_norm": 0.130576491355896, + "learning_rate": 3.89829757401044e-07, + "loss": 0.01, + "step": 111710 + }, + { + "epoch": 32.5050916496945, + "grad_norm": 0.1217254102230072, + "learning_rate": 3.8889403817726596e-07, + "loss": 0.0121, + "step": 111720 + }, + { + "epoch": 32.508001163805645, + "grad_norm": 0.08308164030313492, + "learning_rate": 3.8795942857055957e-07, + "loss": 0.0098, + "step": 111730 + }, + { + "epoch": 32.51091067791679, + "grad_norm": 0.09402818977832794, + "learning_rate": 3.870259286519029e-07, + "loss": 0.0102, + "step": 111740 + }, + { + "epoch": 32.51382019202793, + "grad_norm": 0.13535046577453613, + "learning_rate": 3.860935384921893e-07, + "loss": 0.0093, + "step": 111750 + }, + { + "epoch": 32.516729706139074, + "grad_norm": 0.09611926227807999, + "learning_rate": 3.851622581622305e-07, + "loss": 0.0113, + "step": 111760 + }, + { + "epoch": 32.51963922025022, + "grad_norm": 0.15040569007396698, + "learning_rate": 3.8423208773274654e-07, + "loss": 0.0111, + "step": 111770 + }, + { + "epoch": 32.522548734361365, + "grad_norm": 0.09004873037338257, + "learning_rate": 3.8330302727437924e-07, + "loss": 0.0114, + "step": 111780 + }, + { + "epoch": 32.5254582484725, + "grad_norm": 0.21302726864814758, + "learning_rate": 3.8237507685768724e-07, + "loss": 0.0089, + "step": 111790 + }, + { + "epoch": 32.52836776258365, + "grad_norm": 0.12573082745075226, + "learning_rate": 3.814482365531391e-07, + "loss": 0.0092, + "step": 111800 + }, + { + "epoch": 32.531277276694794, + "grad_norm": 0.11062635481357574, + "learning_rate": 3.805225064311235e-07, + "loss": 0.0107, + "step": 111810 + }, + { + "epoch": 32.53418679080593, + "grad_norm": 0.08884147554636002, + "learning_rate": 3.795978865619426e-07, + "loss": 0.0094, + "step": 111820 + }, + { + "epoch": 32.53709630491708, + "grad_norm": 0.11463578790426254, + "learning_rate": 3.7867437701581854e-07, + "loss": 0.01, + "step": 111830 + }, + { + "epoch": 32.540005819028224, + "grad_norm": 0.19807176291942596, + "learning_rate": 3.7775197786288017e-07, + "loss": 0.0095, + "step": 111840 + }, + { + "epoch": 32.54291533313936, + "grad_norm": 0.12507747113704681, + "learning_rate": 3.7683068917318317e-07, + "loss": 0.0088, + "step": 111850 + }, + { + "epoch": 32.54582484725051, + "grad_norm": 0.11582974344491959, + "learning_rate": 3.759105110166883e-07, + "loss": 0.0105, + "step": 111860 + }, + { + "epoch": 32.54873436136165, + "grad_norm": 0.05433482676744461, + "learning_rate": 3.7499144346328117e-07, + "loss": 0.0088, + "step": 111870 + }, + { + "epoch": 32.5516438754728, + "grad_norm": 0.1670781970024109, + "learning_rate": 3.740734865827561e-07, + "loss": 0.0093, + "step": 111880 + }, + { + "epoch": 32.55455338958394, + "grad_norm": 0.11096125096082687, + "learning_rate": 3.731566404448272e-07, + "loss": 0.01, + "step": 111890 + }, + { + "epoch": 32.55746290369508, + "grad_norm": 0.13282544910907745, + "learning_rate": 3.722409051191239e-07, + "loss": 0.0096, + "step": 111900 + }, + { + "epoch": 32.56037241780623, + "grad_norm": 0.09100012481212616, + "learning_rate": 3.713262806751888e-07, + "loss": 0.0102, + "step": 111910 + }, + { + "epoch": 32.563281931917366, + "grad_norm": 0.12441157549619675, + "learning_rate": 3.704127671824814e-07, + "loss": 0.0112, + "step": 111920 + }, + { + "epoch": 32.56619144602851, + "grad_norm": 0.13893058896064758, + "learning_rate": 3.6950036471037786e-07, + "loss": 0.0118, + "step": 111930 + }, + { + "epoch": 32.56910096013966, + "grad_norm": 0.0781932920217514, + "learning_rate": 3.685890733281677e-07, + "loss": 0.0104, + "step": 111940 + }, + { + "epoch": 32.5720104742508, + "grad_norm": 0.13036972284317017, + "learning_rate": 3.6767889310505895e-07, + "loss": 0.0104, + "step": 111950 + }, + { + "epoch": 32.57491998836194, + "grad_norm": 0.10559894889593124, + "learning_rate": 3.667698241101747e-07, + "loss": 0.0095, + "step": 111960 + }, + { + "epoch": 32.57782950247309, + "grad_norm": 0.10507365316152573, + "learning_rate": 3.6586186641254794e-07, + "loss": 0.0113, + "step": 111970 + }, + { + "epoch": 32.58073901658423, + "grad_norm": 0.10218793153762817, + "learning_rate": 3.6495502008113855e-07, + "loss": 0.0074, + "step": 111980 + }, + { + "epoch": 32.58364853069537, + "grad_norm": 0.07106108218431473, + "learning_rate": 3.640492851848132e-07, + "loss": 0.0123, + "step": 111990 + }, + { + "epoch": 32.586558044806516, + "grad_norm": 0.0677674412727356, + "learning_rate": 3.631446617923534e-07, + "loss": 0.0097, + "step": 112000 + }, + { + "epoch": 32.58946755891766, + "grad_norm": 0.0977654904127121, + "learning_rate": 3.622411499724626e-07, + "loss": 0.0102, + "step": 112010 + }, + { + "epoch": 32.59237707302881, + "grad_norm": 0.09729558229446411, + "learning_rate": 3.613387497937559e-07, + "loss": 0.0115, + "step": 112020 + }, + { + "epoch": 32.595286587139945, + "grad_norm": 0.12168290466070175, + "learning_rate": 3.604374613247652e-07, + "loss": 0.0101, + "step": 112030 + }, + { + "epoch": 32.59819610125109, + "grad_norm": 0.07819133996963501, + "learning_rate": 3.5953728463393565e-07, + "loss": 0.0118, + "step": 112040 + }, + { + "epoch": 32.601105615362236, + "grad_norm": 0.17314469814300537, + "learning_rate": 3.586382197896293e-07, + "loss": 0.0136, + "step": 112050 + }, + { + "epoch": 32.604015129473375, + "grad_norm": 0.14231617748737335, + "learning_rate": 3.5774026686012817e-07, + "loss": 0.009, + "step": 112060 + }, + { + "epoch": 32.60692464358452, + "grad_norm": 0.1770656704902649, + "learning_rate": 3.5684342591362106e-07, + "loss": 0.0113, + "step": 112070 + }, + { + "epoch": 32.609834157695666, + "grad_norm": 0.08865339308977127, + "learning_rate": 3.559476970182185e-07, + "loss": 0.0101, + "step": 112080 + }, + { + "epoch": 32.61274367180681, + "grad_norm": 0.10371583700180054, + "learning_rate": 3.550530802419494e-07, + "loss": 0.0113, + "step": 112090 + }, + { + "epoch": 32.61565318591795, + "grad_norm": 0.07978455722332001, + "learning_rate": 3.541595756527477e-07, + "loss": 0.0098, + "step": 112100 + }, + { + "epoch": 32.618562700029095, + "grad_norm": 0.11771667003631592, + "learning_rate": 3.532671833184742e-07, + "loss": 0.0105, + "step": 112110 + }, + { + "epoch": 32.62147221414024, + "grad_norm": 0.1567104160785675, + "learning_rate": 3.5237590330689793e-07, + "loss": 0.0111, + "step": 112120 + }, + { + "epoch": 32.62438172825138, + "grad_norm": 0.08990521728992462, + "learning_rate": 3.5148573568570485e-07, + "loss": 0.0121, + "step": 112130 + }, + { + "epoch": 32.627291242362524, + "grad_norm": 0.09072469174861908, + "learning_rate": 3.5059668052249916e-07, + "loss": 0.0093, + "step": 112140 + }, + { + "epoch": 32.63020075647367, + "grad_norm": 0.09988945722579956, + "learning_rate": 3.497087378847985e-07, + "loss": 0.0099, + "step": 112150 + }, + { + "epoch": 32.633110270584815, + "grad_norm": 0.10056976228952408, + "learning_rate": 3.4882190784003397e-07, + "loss": 0.0099, + "step": 112160 + }, + { + "epoch": 32.636019784695954, + "grad_norm": 0.08479641377925873, + "learning_rate": 3.4793619045555505e-07, + "loss": 0.0081, + "step": 112170 + }, + { + "epoch": 32.6389292988071, + "grad_norm": 0.10954535752534866, + "learning_rate": 3.4705158579863116e-07, + "loss": 0.0102, + "step": 112180 + }, + { + "epoch": 32.641838812918245, + "grad_norm": 0.15672831237316132, + "learning_rate": 3.461680939364353e-07, + "loss": 0.0109, + "step": 112190 + }, + { + "epoch": 32.64474832702938, + "grad_norm": 0.08521413803100586, + "learning_rate": 3.452857149360672e-07, + "loss": 0.009, + "step": 112200 + }, + { + "epoch": 32.64765784114053, + "grad_norm": 0.11338034272193909, + "learning_rate": 3.444044488645348e-07, + "loss": 0.0095, + "step": 112210 + }, + { + "epoch": 32.650567355251674, + "grad_norm": 0.12237349152565002, + "learning_rate": 3.43524295788768e-07, + "loss": 0.0097, + "step": 112220 + }, + { + "epoch": 32.65347686936282, + "grad_norm": 0.12255679816007614, + "learning_rate": 3.426452557756066e-07, + "loss": 0.0097, + "step": 112230 + }, + { + "epoch": 32.65638638347396, + "grad_norm": 0.0834449976682663, + "learning_rate": 3.417673288918055e-07, + "loss": 0.0101, + "step": 112240 + }, + { + "epoch": 32.6592958975851, + "grad_norm": 0.10250643640756607, + "learning_rate": 3.4089051520404144e-07, + "loss": 0.0108, + "step": 112250 + }, + { + "epoch": 32.66220541169625, + "grad_norm": 0.137957364320755, + "learning_rate": 3.4001481477890284e-07, + "loss": 0.0107, + "step": 112260 + }, + { + "epoch": 32.66511492580739, + "grad_norm": 0.09586115181446075, + "learning_rate": 3.391402276828881e-07, + "loss": 0.0086, + "step": 112270 + }, + { + "epoch": 32.66802443991853, + "grad_norm": 0.07544173300266266, + "learning_rate": 3.382667539824225e-07, + "loss": 0.0104, + "step": 112280 + }, + { + "epoch": 32.67093395402968, + "grad_norm": 0.12036170810461044, + "learning_rate": 3.373943937438362e-07, + "loss": 0.0091, + "step": 112290 + }, + { + "epoch": 32.673843468140824, + "grad_norm": 0.08929306268692017, + "learning_rate": 3.365231470333813e-07, + "loss": 0.0106, + "step": 112300 + }, + { + "epoch": 32.67675298225196, + "grad_norm": 0.16525638103485107, + "learning_rate": 3.356530139172231e-07, + "loss": 0.0094, + "step": 112310 + }, + { + "epoch": 32.67966249636311, + "grad_norm": 0.13282091915607452, + "learning_rate": 3.3478399446144223e-07, + "loss": 0.0094, + "step": 112320 + }, + { + "epoch": 32.68257201047425, + "grad_norm": 0.058094579726457596, + "learning_rate": 3.3391608873203585e-07, + "loss": 0.0084, + "step": 112330 + }, + { + "epoch": 32.68548152458539, + "grad_norm": 0.09103244543075562, + "learning_rate": 3.3304929679491445e-07, + "loss": 0.0094, + "step": 112340 + }, + { + "epoch": 32.68839103869654, + "grad_norm": 0.07913636416196823, + "learning_rate": 3.321836187159039e-07, + "loss": 0.011, + "step": 112350 + }, + { + "epoch": 32.69130055280768, + "grad_norm": 0.09530390799045563, + "learning_rate": 3.313190545607497e-07, + "loss": 0.0096, + "step": 112360 + }, + { + "epoch": 32.69421006691883, + "grad_norm": 0.12582288682460785, + "learning_rate": 3.3045560439510794e-07, + "loss": 0.0087, + "step": 112370 + }, + { + "epoch": 32.697119581029966, + "grad_norm": 0.0548134371638298, + "learning_rate": 3.295932682845526e-07, + "loss": 0.0101, + "step": 112380 + }, + { + "epoch": 32.70002909514111, + "grad_norm": 0.09979178011417389, + "learning_rate": 3.287320462945731e-07, + "loss": 0.0095, + "step": 112390 + }, + { + "epoch": 32.70293860925226, + "grad_norm": 0.10425344109535217, + "learning_rate": 3.278719384905721e-07, + "loss": 0.0121, + "step": 112400 + }, + { + "epoch": 32.705848123363396, + "grad_norm": 0.11888730525970459, + "learning_rate": 3.270129449378689e-07, + "loss": 0.0107, + "step": 112410 + }, + { + "epoch": 32.70875763747454, + "grad_norm": 0.08825445175170898, + "learning_rate": 3.261550657016998e-07, + "loss": 0.0099, + "step": 112420 + }, + { + "epoch": 32.71166715158569, + "grad_norm": 0.08831053227186203, + "learning_rate": 3.25298300847216e-07, + "loss": 0.0096, + "step": 112430 + }, + { + "epoch": 32.71457666569683, + "grad_norm": 0.08775255084037781, + "learning_rate": 3.2444265043947875e-07, + "loss": 0.0096, + "step": 112440 + }, + { + "epoch": 32.71748617980797, + "grad_norm": 0.11735421419143677, + "learning_rate": 3.235881145434744e-07, + "loss": 0.01, + "step": 112450 + }, + { + "epoch": 32.720395693919116, + "grad_norm": 0.11075551062822342, + "learning_rate": 3.2273469322409597e-07, + "loss": 0.0104, + "step": 112460 + }, + { + "epoch": 32.72330520803026, + "grad_norm": 0.11101174354553223, + "learning_rate": 3.218823865461551e-07, + "loss": 0.0102, + "step": 112470 + }, + { + "epoch": 32.7262147221414, + "grad_norm": 0.13197682797908783, + "learning_rate": 3.2103119457438156e-07, + "loss": 0.0102, + "step": 112480 + }, + { + "epoch": 32.729124236252545, + "grad_norm": 0.10370247811079025, + "learning_rate": 3.201811173734154e-07, + "loss": 0.0096, + "step": 112490 + }, + { + "epoch": 32.73203375036369, + "grad_norm": 0.11820723116397858, + "learning_rate": 3.1933215500781486e-07, + "loss": 0.0111, + "step": 112500 + }, + { + "epoch": 32.734943264474836, + "grad_norm": 0.1326059252023697, + "learning_rate": 3.184843075420535e-07, + "loss": 0.0082, + "step": 112510 + }, + { + "epoch": 32.737852778585975, + "grad_norm": 0.09973543137311935, + "learning_rate": 3.176375750405197e-07, + "loss": 0.009, + "step": 112520 + }, + { + "epoch": 32.74076229269712, + "grad_norm": 0.09740941971540451, + "learning_rate": 3.1679195756751543e-07, + "loss": 0.0097, + "step": 112530 + }, + { + "epoch": 32.743671806808266, + "grad_norm": 0.20631635189056396, + "learning_rate": 3.15947455187261e-07, + "loss": 0.0122, + "step": 112540 + }, + { + "epoch": 32.746581320919404, + "grad_norm": 0.09007777273654938, + "learning_rate": 3.1510406796389336e-07, + "loss": 0.0088, + "step": 112550 + }, + { + "epoch": 32.74949083503055, + "grad_norm": 0.12285393476486206, + "learning_rate": 3.1426179596145964e-07, + "loss": 0.0134, + "step": 112560 + }, + { + "epoch": 32.752400349141695, + "grad_norm": 0.12728196382522583, + "learning_rate": 3.1342063924392373e-07, + "loss": 0.0116, + "step": 112570 + }, + { + "epoch": 32.75530986325283, + "grad_norm": 0.13467590510845184, + "learning_rate": 3.125805978751678e-07, + "loss": 0.0088, + "step": 112580 + }, + { + "epoch": 32.75821937736398, + "grad_norm": 0.09037640690803528, + "learning_rate": 3.117416719189875e-07, + "loss": 0.0085, + "step": 112590 + }, + { + "epoch": 32.761128891475124, + "grad_norm": 0.3208988904953003, + "learning_rate": 3.1090386143909186e-07, + "loss": 0.009, + "step": 112600 + }, + { + "epoch": 32.76403840558627, + "grad_norm": 0.10400837659835815, + "learning_rate": 3.1006716649910835e-07, + "loss": 0.0099, + "step": 112610 + }, + { + "epoch": 32.76694791969741, + "grad_norm": 0.11001506447792053, + "learning_rate": 3.092315871625795e-07, + "loss": 0.0108, + "step": 112620 + }, + { + "epoch": 32.76985743380855, + "grad_norm": 0.08171840757131577, + "learning_rate": 3.0839712349296113e-07, + "loss": 0.0077, + "step": 112630 + }, + { + "epoch": 32.7727669479197, + "grad_norm": 0.07080948352813721, + "learning_rate": 3.0756377555362425e-07, + "loss": 0.007, + "step": 112640 + }, + { + "epoch": 32.77567646203084, + "grad_norm": 0.08289693295955658, + "learning_rate": 3.0673154340785656e-07, + "loss": 0.0102, + "step": 112650 + }, + { + "epoch": 32.77858597614198, + "grad_norm": 0.12500505149364471, + "learning_rate": 3.0590042711886257e-07, + "loss": 0.009, + "step": 112660 + }, + { + "epoch": 32.78149549025313, + "grad_norm": 0.09072291105985641, + "learning_rate": 3.050704267497584e-07, + "loss": 0.0105, + "step": 112670 + }, + { + "epoch": 32.784405004364274, + "grad_norm": 0.11598008871078491, + "learning_rate": 3.042415423635736e-07, + "loss": 0.0093, + "step": 112680 + }, + { + "epoch": 32.78731451847541, + "grad_norm": 0.0889853984117508, + "learning_rate": 3.034137740232645e-07, + "loss": 0.0105, + "step": 112690 + }, + { + "epoch": 32.79022403258656, + "grad_norm": 0.09205038100481033, + "learning_rate": 3.0258712179168924e-07, + "loss": 0.0081, + "step": 112700 + }, + { + "epoch": 32.7931335466977, + "grad_norm": 0.15595465898513794, + "learning_rate": 3.017615857316258e-07, + "loss": 0.0102, + "step": 112710 + }, + { + "epoch": 32.79604306080884, + "grad_norm": 0.10932721942663193, + "learning_rate": 3.009371659057708e-07, + "loss": 0.0111, + "step": 112720 + }, + { + "epoch": 32.79895257491999, + "grad_norm": 0.14322657883167267, + "learning_rate": 3.001138623767341e-07, + "loss": 0.0113, + "step": 112730 + }, + { + "epoch": 32.80186208903113, + "grad_norm": 0.11328393965959549, + "learning_rate": 2.9929167520703737e-07, + "loss": 0.0103, + "step": 112740 + }, + { + "epoch": 32.80477160314228, + "grad_norm": 0.09843549877405167, + "learning_rate": 2.984706044591223e-07, + "loss": 0.0121, + "step": 112750 + }, + { + "epoch": 32.807681117253416, + "grad_norm": 0.08228198438882828, + "learning_rate": 2.976506501953424e-07, + "loss": 0.0097, + "step": 112760 + }, + { + "epoch": 32.81059063136456, + "grad_norm": 0.12125616520643234, + "learning_rate": 2.968318124779695e-07, + "loss": 0.0094, + "step": 112770 + }, + { + "epoch": 32.81350014547571, + "grad_norm": 0.1638822704553604, + "learning_rate": 2.9601409136918723e-07, + "loss": 0.0091, + "step": 112780 + }, + { + "epoch": 32.816409659586846, + "grad_norm": 0.1122971773147583, + "learning_rate": 2.951974869310958e-07, + "loss": 0.0094, + "step": 112790 + }, + { + "epoch": 32.81931917369799, + "grad_norm": 0.08341007679700851, + "learning_rate": 2.943819992257141e-07, + "loss": 0.0099, + "step": 112800 + }, + { + "epoch": 32.82222868780914, + "grad_norm": 0.0893196314573288, + "learning_rate": 2.9356762831497076e-07, + "loss": 0.0092, + "step": 112810 + }, + { + "epoch": 32.82513820192028, + "grad_norm": 0.09626497328281403, + "learning_rate": 2.927543742607114e-07, + "loss": 0.0102, + "step": 112820 + }, + { + "epoch": 32.82804771603142, + "grad_norm": 0.09905099123716354, + "learning_rate": 2.9194223712469825e-07, + "loss": 0.0098, + "step": 112830 + }, + { + "epoch": 32.830957230142566, + "grad_norm": 0.10025909543037415, + "learning_rate": 2.9113121696860537e-07, + "loss": 0.0112, + "step": 112840 + }, + { + "epoch": 32.83386674425371, + "grad_norm": 0.105056531727314, + "learning_rate": 2.9032131385403004e-07, + "loss": 0.0107, + "step": 112850 + }, + { + "epoch": 32.83677625836485, + "grad_norm": 0.08639685064554214, + "learning_rate": 2.8951252784247317e-07, + "loss": 0.0102, + "step": 112860 + }, + { + "epoch": 32.839685772475995, + "grad_norm": 0.08905529230833054, + "learning_rate": 2.8870485899535893e-07, + "loss": 0.0128, + "step": 112870 + }, + { + "epoch": 32.84259528658714, + "grad_norm": 0.11838354170322418, + "learning_rate": 2.878983073740249e-07, + "loss": 0.0095, + "step": 112880 + }, + { + "epoch": 32.845504800698286, + "grad_norm": 0.11335836350917816, + "learning_rate": 2.8709287303972373e-07, + "loss": 0.0087, + "step": 112890 + }, + { + "epoch": 32.848414314809425, + "grad_norm": 0.12598459422588348, + "learning_rate": 2.862885560536216e-07, + "loss": 0.0092, + "step": 112900 + }, + { + "epoch": 32.85132382892057, + "grad_norm": 0.1186632290482521, + "learning_rate": 2.854853564768012e-07, + "loss": 0.0109, + "step": 112910 + }, + { + "epoch": 32.854233343031716, + "grad_norm": 0.08748916536569595, + "learning_rate": 2.846832743702621e-07, + "loss": 0.0091, + "step": 112920 + }, + { + "epoch": 32.857142857142854, + "grad_norm": 0.08060594648122787, + "learning_rate": 2.838823097949156e-07, + "loss": 0.011, + "step": 112930 + }, + { + "epoch": 32.860052371254, + "grad_norm": 0.14499728381633759, + "learning_rate": 2.830824628115863e-07, + "loss": 0.0108, + "step": 112940 + }, + { + "epoch": 32.862961885365145, + "grad_norm": 0.11563021689653397, + "learning_rate": 2.822837334810241e-07, + "loss": 0.0103, + "step": 112950 + }, + { + "epoch": 32.86587139947629, + "grad_norm": 0.08188991993665695, + "learning_rate": 2.814861218638837e-07, + "loss": 0.0105, + "step": 112960 + }, + { + "epoch": 32.86878091358743, + "grad_norm": 0.12633419036865234, + "learning_rate": 2.8068962802073995e-07, + "loss": 0.0083, + "step": 112970 + }, + { + "epoch": 32.871690427698574, + "grad_norm": 0.09361465275287628, + "learning_rate": 2.7989425201207453e-07, + "loss": 0.0122, + "step": 112980 + }, + { + "epoch": 32.87459994180972, + "grad_norm": 0.195834219455719, + "learning_rate": 2.7909999389830076e-07, + "loss": 0.0098, + "step": 112990 + }, + { + "epoch": 32.87750945592086, + "grad_norm": 0.10482145845890045, + "learning_rate": 2.7830685373973375e-07, + "loss": 0.0103, + "step": 113000 + }, + { + "epoch": 32.880418970032004, + "grad_norm": 0.10053907334804535, + "learning_rate": 2.7751483159660364e-07, + "loss": 0.0099, + "step": 113010 + }, + { + "epoch": 32.88332848414315, + "grad_norm": 0.12031325697898865, + "learning_rate": 2.767239275290623e-07, + "loss": 0.0101, + "step": 113020 + }, + { + "epoch": 32.886237998254295, + "grad_norm": 0.18275347352027893, + "learning_rate": 2.75934141597175e-07, + "loss": 0.0104, + "step": 113030 + }, + { + "epoch": 32.88914751236543, + "grad_norm": 0.0899670273065567, + "learning_rate": 2.751454738609205e-07, + "loss": 0.0101, + "step": 113040 + }, + { + "epoch": 32.89205702647658, + "grad_norm": 0.11232806742191315, + "learning_rate": 2.7435792438018747e-07, + "loss": 0.0093, + "step": 113050 + }, + { + "epoch": 32.894966540587724, + "grad_norm": 0.11384978145360947, + "learning_rate": 2.7357149321479315e-07, + "loss": 0.0108, + "step": 113060 + }, + { + "epoch": 32.89787605469886, + "grad_norm": 0.09309344738721848, + "learning_rate": 2.7278618042445637e-07, + "loss": 0.0111, + "step": 113070 + }, + { + "epoch": 32.90078556881001, + "grad_norm": 0.09238001704216003, + "learning_rate": 2.720019860688178e-07, + "loss": 0.0081, + "step": 113080 + }, + { + "epoch": 32.90369508292115, + "grad_norm": 0.21325984597206116, + "learning_rate": 2.7121891020743305e-07, + "loss": 0.0096, + "step": 113090 + }, + { + "epoch": 32.9066045970323, + "grad_norm": 0.10024705529212952, + "learning_rate": 2.704369528997697e-07, + "loss": 0.0105, + "step": 113100 + }, + { + "epoch": 32.90951411114344, + "grad_norm": 0.07796116918325424, + "learning_rate": 2.696561142052134e-07, + "loss": 0.0095, + "step": 113110 + }, + { + "epoch": 32.91242362525458, + "grad_norm": 0.15487992763519287, + "learning_rate": 2.688763941830635e-07, + "loss": 0.0097, + "step": 113120 + }, + { + "epoch": 32.91533313936573, + "grad_norm": 0.10542180389165878, + "learning_rate": 2.6809779289253435e-07, + "loss": 0.011, + "step": 113130 + }, + { + "epoch": 32.91824265347687, + "grad_norm": 0.12981706857681274, + "learning_rate": 2.6732031039275684e-07, + "loss": 0.0127, + "step": 113140 + }, + { + "epoch": 32.92115216758801, + "grad_norm": 0.12542155385017395, + "learning_rate": 2.6654394674277216e-07, + "loss": 0.0126, + "step": 113150 + }, + { + "epoch": 32.92406168169916, + "grad_norm": 0.13123798370361328, + "learning_rate": 2.657687020015431e-07, + "loss": 0.0115, + "step": 113160 + }, + { + "epoch": 32.9269711958103, + "grad_norm": 0.15597540140151978, + "learning_rate": 2.6499457622794596e-07, + "loss": 0.0088, + "step": 113170 + }, + { + "epoch": 32.92988070992144, + "grad_norm": 0.14452043175697327, + "learning_rate": 2.6422156948076536e-07, + "loss": 0.0107, + "step": 113180 + }, + { + "epoch": 32.93279022403259, + "grad_norm": 0.10465177148580551, + "learning_rate": 2.6344968181871097e-07, + "loss": 0.0104, + "step": 113190 + }, + { + "epoch": 32.93569973814373, + "grad_norm": 0.07207321375608444, + "learning_rate": 2.626789133003993e-07, + "loss": 0.0106, + "step": 113200 + }, + { + "epoch": 32.93860925225487, + "grad_norm": 0.1190493106842041, + "learning_rate": 2.619092639843651e-07, + "loss": 0.0102, + "step": 113210 + }, + { + "epoch": 32.941518766366016, + "grad_norm": 0.10432503372430801, + "learning_rate": 2.6114073392906166e-07, + "loss": 0.0107, + "step": 113220 + }, + { + "epoch": 32.94442828047716, + "grad_norm": 0.10363354533910751, + "learning_rate": 2.603733231928507e-07, + "loss": 0.0081, + "step": 113230 + }, + { + "epoch": 32.94733779458831, + "grad_norm": 0.10553332418203354, + "learning_rate": 2.5960703183401215e-07, + "loss": 0.0105, + "step": 113240 + }, + { + "epoch": 32.950247308699446, + "grad_norm": 0.10801335424184799, + "learning_rate": 2.588418599107428e-07, + "loss": 0.0094, + "step": 113250 + }, + { + "epoch": 32.95315682281059, + "grad_norm": 0.1387072503566742, + "learning_rate": 2.5807780748114963e-07, + "loss": 0.0088, + "step": 113260 + }, + { + "epoch": 32.95606633692174, + "grad_norm": 0.07610314339399338, + "learning_rate": 2.573148746032594e-07, + "loss": 0.0099, + "step": 113270 + }, + { + "epoch": 32.958975851032875, + "grad_norm": 0.1130502000451088, + "learning_rate": 2.565530613350109e-07, + "loss": 0.0104, + "step": 113280 + }, + { + "epoch": 32.96188536514402, + "grad_norm": 0.06077751889824867, + "learning_rate": 2.557923677342594e-07, + "loss": 0.01, + "step": 113290 + }, + { + "epoch": 32.964794879255166, + "grad_norm": 0.06468594819307327, + "learning_rate": 2.550327938587754e-07, + "loss": 0.0087, + "step": 113300 + }, + { + "epoch": 32.967704393366304, + "grad_norm": 0.12071669846773148, + "learning_rate": 2.542743397662395e-07, + "loss": 0.0095, + "step": 113310 + }, + { + "epoch": 32.97061390747745, + "grad_norm": 0.0885557234287262, + "learning_rate": 2.535170055142555e-07, + "loss": 0.009, + "step": 113320 + }, + { + "epoch": 32.973523421588595, + "grad_norm": 0.1973399817943573, + "learning_rate": 2.5276079116033747e-07, + "loss": 0.0129, + "step": 113330 + }, + { + "epoch": 32.97643293569974, + "grad_norm": 0.08399847149848938, + "learning_rate": 2.5200569676191444e-07, + "loss": 0.0098, + "step": 113340 + }, + { + "epoch": 32.97934244981088, + "grad_norm": 0.11781756579875946, + "learning_rate": 2.512517223763272e-07, + "loss": 0.0115, + "step": 113350 + }, + { + "epoch": 32.982251963922025, + "grad_norm": 0.08295045793056488, + "learning_rate": 2.5049886806084163e-07, + "loss": 0.0096, + "step": 113360 + }, + { + "epoch": 32.98516147803317, + "grad_norm": 0.14667880535125732, + "learning_rate": 2.497471338726287e-07, + "loss": 0.0111, + "step": 113370 + }, + { + "epoch": 32.98807099214431, + "grad_norm": 0.07683733105659485, + "learning_rate": 2.48996519868776e-07, + "loss": 0.0118, + "step": 113380 + }, + { + "epoch": 32.990980506255454, + "grad_norm": 0.1357795000076294, + "learning_rate": 2.482470261062897e-07, + "loss": 0.0106, + "step": 113390 + }, + { + "epoch": 32.9938900203666, + "grad_norm": 0.1141398698091507, + "learning_rate": 2.4749865264209083e-07, + "loss": 0.0102, + "step": 113400 + }, + { + "epoch": 32.996799534477745, + "grad_norm": 0.12796789407730103, + "learning_rate": 2.467513995330106e-07, + "loss": 0.0127, + "step": 113410 + }, + { + "epoch": 32.99970904858888, + "grad_norm": 0.11440057307481766, + "learning_rate": 2.4600526683579705e-07, + "loss": 0.0107, + "step": 113420 + }, + { + "epoch": 33.00261856270003, + "grad_norm": 0.11050476878881454, + "learning_rate": 2.4526025460711476e-07, + "loss": 0.0109, + "step": 113430 + }, + { + "epoch": 33.005528076811174, + "grad_norm": 0.08690178394317627, + "learning_rate": 2.4451636290354683e-07, + "loss": 0.0132, + "step": 113440 + }, + { + "epoch": 33.00843759092231, + "grad_norm": 0.057297322899103165, + "learning_rate": 2.4377359178157977e-07, + "loss": 0.009, + "step": 113450 + }, + { + "epoch": 33.01134710503346, + "grad_norm": 0.15307940542697906, + "learning_rate": 2.4303194129762676e-07, + "loss": 0.0095, + "step": 113460 + }, + { + "epoch": 33.014256619144604, + "grad_norm": 0.08126943558454514, + "learning_rate": 2.422914115080127e-07, + "loss": 0.0091, + "step": 113470 + }, + { + "epoch": 33.01716613325575, + "grad_norm": 0.08572998642921448, + "learning_rate": 2.41552002468971e-07, + "loss": 0.0092, + "step": 113480 + }, + { + "epoch": 33.02007564736689, + "grad_norm": 0.12511026859283447, + "learning_rate": 2.4081371423665843e-07, + "loss": 0.0118, + "step": 113490 + }, + { + "epoch": 33.02298516147803, + "grad_norm": 0.1469980776309967, + "learning_rate": 2.4007654686714175e-07, + "loss": 0.0105, + "step": 113500 + }, + { + "epoch": 33.02589467558918, + "grad_norm": 0.09868685156106949, + "learning_rate": 2.3934050041640456e-07, + "loss": 0.0109, + "step": 113510 + }, + { + "epoch": 33.02880418970032, + "grad_norm": 0.09058651328086853, + "learning_rate": 2.386055749403454e-07, + "loss": 0.0125, + "step": 113520 + }, + { + "epoch": 33.03171370381146, + "grad_norm": 0.11022749543190002, + "learning_rate": 2.3787177049477472e-07, + "loss": 0.0097, + "step": 113530 + }, + { + "epoch": 33.03462321792261, + "grad_norm": 0.08677084743976593, + "learning_rate": 2.371390871354212e-07, + "loss": 0.01, + "step": 113540 + }, + { + "epoch": 33.03753273203375, + "grad_norm": 0.09893183410167694, + "learning_rate": 2.364075249179287e-07, + "loss": 0.0117, + "step": 113550 + }, + { + "epoch": 33.04044224614489, + "grad_norm": 0.15411795675754547, + "learning_rate": 2.3567708389785282e-07, + "loss": 0.0097, + "step": 113560 + }, + { + "epoch": 33.04335176025604, + "grad_norm": 0.1190224140882492, + "learning_rate": 2.3494776413066744e-07, + "loss": 0.0115, + "step": 113570 + }, + { + "epoch": 33.04626127436718, + "grad_norm": 0.1034713014960289, + "learning_rate": 2.3421956567175663e-07, + "loss": 0.0079, + "step": 113580 + }, + { + "epoch": 33.04917078847832, + "grad_norm": 0.11078616976737976, + "learning_rate": 2.3349248857642612e-07, + "loss": 0.01, + "step": 113590 + }, + { + "epoch": 33.05208030258947, + "grad_norm": 0.09900974482297897, + "learning_rate": 2.3276653289989014e-07, + "loss": 0.0122, + "step": 113600 + }, + { + "epoch": 33.05498981670061, + "grad_norm": 0.09511622041463852, + "learning_rate": 2.3204169869727953e-07, + "loss": 0.0138, + "step": 113610 + }, + { + "epoch": 33.05789933081176, + "grad_norm": 0.08155988156795502, + "learning_rate": 2.3131798602364197e-07, + "loss": 0.0101, + "step": 113620 + }, + { + "epoch": 33.060808844922896, + "grad_norm": 0.1755489557981491, + "learning_rate": 2.305953949339401e-07, + "loss": 0.012, + "step": 113630 + }, + { + "epoch": 33.06371835903404, + "grad_norm": 0.07273174077272415, + "learning_rate": 2.2987392548304841e-07, + "loss": 0.0109, + "step": 113640 + }, + { + "epoch": 33.06662787314519, + "grad_norm": 0.1844385266304016, + "learning_rate": 2.2915357772575473e-07, + "loss": 0.0105, + "step": 113650 + }, + { + "epoch": 33.069537387256325, + "grad_norm": 0.06846442818641663, + "learning_rate": 2.2843435171677029e-07, + "loss": 0.0099, + "step": 113660 + }, + { + "epoch": 33.07244690136747, + "grad_norm": 0.10699552297592163, + "learning_rate": 2.2771624751071308e-07, + "loss": 0.0104, + "step": 113670 + }, + { + "epoch": 33.075356415478616, + "grad_norm": 0.1048256903886795, + "learning_rate": 2.2699926516211778e-07, + "loss": 0.0128, + "step": 113680 + }, + { + "epoch": 33.07826592958976, + "grad_norm": 0.22278115153312683, + "learning_rate": 2.262834047254342e-07, + "loss": 0.0103, + "step": 113690 + }, + { + "epoch": 33.0811754437009, + "grad_norm": 0.0973890870809555, + "learning_rate": 2.2556866625502881e-07, + "loss": 0.0109, + "step": 113700 + }, + { + "epoch": 33.084084957812046, + "grad_norm": 0.12602873146533966, + "learning_rate": 2.2485504980518156e-07, + "loss": 0.0097, + "step": 113710 + }, + { + "epoch": 33.08699447192319, + "grad_norm": 0.11943813413381577, + "learning_rate": 2.241425554300841e-07, + "loss": 0.0106, + "step": 113720 + }, + { + "epoch": 33.08990398603433, + "grad_norm": 0.07241042703390121, + "learning_rate": 2.2343118318384648e-07, + "loss": 0.0105, + "step": 113730 + }, + { + "epoch": 33.092813500145475, + "grad_norm": 0.09410500526428223, + "learning_rate": 2.2272093312049546e-07, + "loss": 0.0106, + "step": 113740 + }, + { + "epoch": 33.09572301425662, + "grad_norm": 0.11516346037387848, + "learning_rate": 2.2201180529396792e-07, + "loss": 0.0081, + "step": 113750 + }, + { + "epoch": 33.098632528367766, + "grad_norm": 0.0879523828625679, + "learning_rate": 2.2130379975811576e-07, + "loss": 0.0102, + "step": 113760 + }, + { + "epoch": 33.101542042478904, + "grad_norm": 0.10143303871154785, + "learning_rate": 2.20596916566711e-07, + "loss": 0.0114, + "step": 113770 + }, + { + "epoch": 33.10445155659005, + "grad_norm": 0.17294614017009735, + "learning_rate": 2.1989115577343566e-07, + "loss": 0.0118, + "step": 113780 + }, + { + "epoch": 33.107361070701195, + "grad_norm": 0.13975852727890015, + "learning_rate": 2.1918651743188522e-07, + "loss": 0.0114, + "step": 113790 + }, + { + "epoch": 33.11027058481233, + "grad_norm": 0.07805381715297699, + "learning_rate": 2.1848300159557355e-07, + "loss": 0.0106, + "step": 113800 + }, + { + "epoch": 33.11318009892348, + "grad_norm": 0.11508242040872574, + "learning_rate": 2.1778060831792957e-07, + "loss": 0.0096, + "step": 113810 + }, + { + "epoch": 33.116089613034625, + "grad_norm": 0.1194833368062973, + "learning_rate": 2.1707933765229392e-07, + "loss": 0.0125, + "step": 113820 + }, + { + "epoch": 33.11899912714577, + "grad_norm": 0.12383440881967545, + "learning_rate": 2.1637918965192405e-07, + "loss": 0.0094, + "step": 113830 + }, + { + "epoch": 33.12190864125691, + "grad_norm": 0.06998740136623383, + "learning_rate": 2.1568016436998904e-07, + "loss": 0.0106, + "step": 113840 + }, + { + "epoch": 33.124818155368054, + "grad_norm": 0.1258663535118103, + "learning_rate": 2.1498226185957981e-07, + "loss": 0.0086, + "step": 113850 + }, + { + "epoch": 33.1277276694792, + "grad_norm": 0.07482526451349258, + "learning_rate": 2.1428548217369558e-07, + "loss": 0.009, + "step": 113860 + }, + { + "epoch": 33.13063718359034, + "grad_norm": 0.10394585132598877, + "learning_rate": 2.1358982536524907e-07, + "loss": 0.009, + "step": 113870 + }, + { + "epoch": 33.13354669770148, + "grad_norm": 0.07794881612062454, + "learning_rate": 2.1289529148707632e-07, + "loss": 0.0103, + "step": 113880 + }, + { + "epoch": 33.13645621181263, + "grad_norm": 0.09218138456344604, + "learning_rate": 2.1220188059191848e-07, + "loss": 0.0092, + "step": 113890 + }, + { + "epoch": 33.139365725923774, + "grad_norm": 0.07111234217882156, + "learning_rate": 2.1150959273243674e-07, + "loss": 0.0108, + "step": 113900 + }, + { + "epoch": 33.14227524003491, + "grad_norm": 0.09048547595739365, + "learning_rate": 2.1081842796120575e-07, + "loss": 0.0119, + "step": 113910 + }, + { + "epoch": 33.14518475414606, + "grad_norm": 0.09072431176900864, + "learning_rate": 2.1012838633071518e-07, + "loss": 0.0112, + "step": 113920 + }, + { + "epoch": 33.1480942682572, + "grad_norm": 0.08520019799470901, + "learning_rate": 2.0943946789336975e-07, + "loss": 0.0102, + "step": 113930 + }, + { + "epoch": 33.15100378236834, + "grad_norm": 0.11448948085308075, + "learning_rate": 2.087516727014893e-07, + "loss": 0.0103, + "step": 113940 + }, + { + "epoch": 33.15391329647949, + "grad_norm": 0.12774047255516052, + "learning_rate": 2.080650008073004e-07, + "loss": 0.012, + "step": 113950 + }, + { + "epoch": 33.15682281059063, + "grad_norm": 0.21321485936641693, + "learning_rate": 2.073794522629613e-07, + "loss": 0.009, + "step": 113960 + }, + { + "epoch": 33.15973232470178, + "grad_norm": 0.10949594527482986, + "learning_rate": 2.066950271205287e-07, + "loss": 0.0106, + "step": 113970 + }, + { + "epoch": 33.16264183881292, + "grad_norm": 0.08765363693237305, + "learning_rate": 2.0601172543197933e-07, + "loss": 0.0123, + "step": 113980 + }, + { + "epoch": 33.16555135292406, + "grad_norm": 0.16529372334480286, + "learning_rate": 2.053295472492084e-07, + "loss": 0.0097, + "step": 113990 + }, + { + "epoch": 33.16846086703521, + "grad_norm": 0.11635785549879074, + "learning_rate": 2.046484926240244e-07, + "loss": 0.0096, + "step": 114000 + }, + { + "epoch": 33.171370381146346, + "grad_norm": 0.10841956734657288, + "learning_rate": 2.0396856160814603e-07, + "loss": 0.0103, + "step": 114010 + }, + { + "epoch": 33.17427989525749, + "grad_norm": 0.09105563908815384, + "learning_rate": 2.0328975425320694e-07, + "loss": 0.0108, + "step": 114020 + }, + { + "epoch": 33.17718940936864, + "grad_norm": 0.0722188949584961, + "learning_rate": 2.0261207061076425e-07, + "loss": 0.0104, + "step": 114030 + }, + { + "epoch": 33.180098923479775, + "grad_norm": 0.08599385619163513, + "learning_rate": 2.019355107322801e-07, + "loss": 0.0104, + "step": 114040 + }, + { + "epoch": 33.18300843759092, + "grad_norm": 0.1578143835067749, + "learning_rate": 2.0126007466913344e-07, + "loss": 0.0104, + "step": 114050 + }, + { + "epoch": 33.185917951702066, + "grad_norm": 0.09752547740936279, + "learning_rate": 2.005857624726215e-07, + "loss": 0.0132, + "step": 114060 + }, + { + "epoch": 33.18882746581321, + "grad_norm": 0.07992703467607498, + "learning_rate": 1.9991257419395502e-07, + "loss": 0.0102, + "step": 114070 + }, + { + "epoch": 33.19173697992435, + "grad_norm": 0.10856842994689941, + "learning_rate": 1.992405098842548e-07, + "loss": 0.0112, + "step": 114080 + }, + { + "epoch": 33.194646494035496, + "grad_norm": 0.09614811092615128, + "learning_rate": 1.9856956959456163e-07, + "loss": 0.0104, + "step": 114090 + }, + { + "epoch": 33.19755600814664, + "grad_norm": 0.1142117828130722, + "learning_rate": 1.9789975337582812e-07, + "loss": 0.0106, + "step": 114100 + }, + { + "epoch": 33.20046552225778, + "grad_norm": 0.13116233050823212, + "learning_rate": 1.972310612789252e-07, + "loss": 0.0105, + "step": 114110 + }, + { + "epoch": 33.203375036368925, + "grad_norm": 0.12359324842691422, + "learning_rate": 1.965634933546323e-07, + "loss": 0.0114, + "step": 114120 + }, + { + "epoch": 33.20628455048007, + "grad_norm": 0.10956642031669617, + "learning_rate": 1.958970496536472e-07, + "loss": 0.0107, + "step": 114130 + }, + { + "epoch": 33.209194064591216, + "grad_norm": 0.06521038711071014, + "learning_rate": 1.952317302265827e-07, + "loss": 0.01, + "step": 114140 + }, + { + "epoch": 33.212103578702354, + "grad_norm": 0.14233016967773438, + "learning_rate": 1.9456753512396508e-07, + "loss": 0.0086, + "step": 114150 + }, + { + "epoch": 33.2150130928135, + "grad_norm": 0.09640666842460632, + "learning_rate": 1.9390446439623733e-07, + "loss": 0.0111, + "step": 114160 + }, + { + "epoch": 33.217922606924645, + "grad_norm": 0.11295308917760849, + "learning_rate": 1.932425180937525e-07, + "loss": 0.0124, + "step": 114170 + }, + { + "epoch": 33.220832121035784, + "grad_norm": 0.09635929763317108, + "learning_rate": 1.92581696266782e-07, + "loss": 0.009, + "step": 114180 + }, + { + "epoch": 33.22374163514693, + "grad_norm": 0.07521814107894897, + "learning_rate": 1.9192199896551243e-07, + "loss": 0.0104, + "step": 114190 + }, + { + "epoch": 33.226651149258075, + "grad_norm": 0.09553618729114532, + "learning_rate": 1.91263426240042e-07, + "loss": 0.0108, + "step": 114200 + }, + { + "epoch": 33.22956066336922, + "grad_norm": 0.12144821137189865, + "learning_rate": 1.9060597814038404e-07, + "loss": 0.0103, + "step": 114210 + }, + { + "epoch": 33.23247017748036, + "grad_norm": 0.15452967584133148, + "learning_rate": 1.8994965471646863e-07, + "loss": 0.0105, + "step": 114220 + }, + { + "epoch": 33.235379691591504, + "grad_norm": 0.10495667904615402, + "learning_rate": 1.892944560181409e-07, + "loss": 0.0105, + "step": 114230 + }, + { + "epoch": 33.23828920570265, + "grad_norm": 0.11087114363908768, + "learning_rate": 1.8864038209515598e-07, + "loss": 0.0098, + "step": 114240 + }, + { + "epoch": 33.24119871981379, + "grad_norm": 0.12453994899988174, + "learning_rate": 1.8798743299718757e-07, + "loss": 0.0108, + "step": 114250 + }, + { + "epoch": 33.24410823392493, + "grad_norm": 0.07090442627668381, + "learning_rate": 1.873356087738226e-07, + "loss": 0.0084, + "step": 114260 + }, + { + "epoch": 33.24701774803608, + "grad_norm": 0.06383980065584183, + "learning_rate": 1.8668490947456317e-07, + "loss": 0.0102, + "step": 114270 + }, + { + "epoch": 33.249927262147224, + "grad_norm": 0.10614398121833801, + "learning_rate": 1.8603533514882642e-07, + "loss": 0.0109, + "step": 114280 + }, + { + "epoch": 33.25283677625836, + "grad_norm": 0.17299754917621613, + "learning_rate": 1.8538688584594123e-07, + "loss": 0.0097, + "step": 114290 + }, + { + "epoch": 33.25574629036951, + "grad_norm": 0.08757849782705307, + "learning_rate": 1.8473956161515481e-07, + "loss": 0.0098, + "step": 114300 + }, + { + "epoch": 33.258655804480654, + "grad_norm": 0.14491645991802216, + "learning_rate": 1.8409336250562626e-07, + "loss": 0.0123, + "step": 114310 + }, + { + "epoch": 33.26156531859179, + "grad_norm": 0.0942610651254654, + "learning_rate": 1.834482885664296e-07, + "loss": 0.0097, + "step": 114320 + }, + { + "epoch": 33.26447483270294, + "grad_norm": 0.15184688568115234, + "learning_rate": 1.82804339846554e-07, + "loss": 0.0113, + "step": 114330 + }, + { + "epoch": 33.26738434681408, + "grad_norm": 0.08741419017314911, + "learning_rate": 1.8216151639490696e-07, + "loss": 0.0091, + "step": 114340 + }, + { + "epoch": 33.27029386092523, + "grad_norm": 0.09582412242889404, + "learning_rate": 1.8151981826030117e-07, + "loss": 0.0093, + "step": 114350 + }, + { + "epoch": 33.27320337503637, + "grad_norm": 0.07632077485322952, + "learning_rate": 1.8087924549147094e-07, + "loss": 0.0084, + "step": 114360 + }, + { + "epoch": 33.27611288914751, + "grad_norm": 0.12936431169509888, + "learning_rate": 1.802397981370657e-07, + "loss": 0.0113, + "step": 114370 + }, + { + "epoch": 33.27902240325866, + "grad_norm": 0.0838799923658371, + "learning_rate": 1.796014762456466e-07, + "loss": 0.01, + "step": 114380 + }, + { + "epoch": 33.281931917369796, + "grad_norm": 0.12660086154937744, + "learning_rate": 1.789642798656882e-07, + "loss": 0.0112, + "step": 114390 + }, + { + "epoch": 33.28484143148094, + "grad_norm": 0.09279454499483109, + "learning_rate": 1.7832820904558178e-07, + "loss": 0.0103, + "step": 114400 + }, + { + "epoch": 33.28775094559209, + "grad_norm": 0.10074537992477417, + "learning_rate": 1.7769326383363538e-07, + "loss": 0.0122, + "step": 114410 + }, + { + "epoch": 33.29066045970323, + "grad_norm": 0.13696125149726868, + "learning_rate": 1.7705944427806543e-07, + "loss": 0.0098, + "step": 114420 + }, + { + "epoch": 33.29356997381437, + "grad_norm": 0.1050359383225441, + "learning_rate": 1.764267504270084e-07, + "loss": 0.0106, + "step": 114430 + }, + { + "epoch": 33.29647948792552, + "grad_norm": 0.17575575411319733, + "learning_rate": 1.7579518232851256e-07, + "loss": 0.0108, + "step": 114440 + }, + { + "epoch": 33.29938900203666, + "grad_norm": 0.09463081508874893, + "learning_rate": 1.7516474003054118e-07, + "loss": 0.0127, + "step": 114450 + }, + { + "epoch": 33.3022985161478, + "grad_norm": 0.12295771390199661, + "learning_rate": 1.7453542358097097e-07, + "loss": 0.0094, + "step": 114460 + }, + { + "epoch": 33.305208030258946, + "grad_norm": 0.08456528931856155, + "learning_rate": 1.7390723302759704e-07, + "loss": 0.0107, + "step": 114470 + }, + { + "epoch": 33.30811754437009, + "grad_norm": 0.10078727453947067, + "learning_rate": 1.7328016841812622e-07, + "loss": 0.0111, + "step": 114480 + }, + { + "epoch": 33.31102705848124, + "grad_norm": 0.09858165681362152, + "learning_rate": 1.7265422980017709e-07, + "loss": 0.0095, + "step": 114490 + }, + { + "epoch": 33.313936572592375, + "grad_norm": 0.09099265933036804, + "learning_rate": 1.720294172212883e-07, + "loss": 0.011, + "step": 114500 + }, + { + "epoch": 33.31684608670352, + "grad_norm": 0.10051220655441284, + "learning_rate": 1.7140573072890853e-07, + "loss": 0.0088, + "step": 114510 + }, + { + "epoch": 33.319755600814666, + "grad_norm": 0.13587065041065216, + "learning_rate": 1.7078317037040326e-07, + "loss": 0.0135, + "step": 114520 + }, + { + "epoch": 33.322665114925805, + "grad_norm": 0.09997562319040298, + "learning_rate": 1.7016173619305298e-07, + "loss": 0.011, + "step": 114530 + }, + { + "epoch": 33.32557462903695, + "grad_norm": 0.09763843566179276, + "learning_rate": 1.6954142824404995e-07, + "loss": 0.0109, + "step": 114540 + }, + { + "epoch": 33.328484143148096, + "grad_norm": 0.16648532450199127, + "learning_rate": 1.689222465705015e-07, + "loss": 0.0093, + "step": 114550 + }, + { + "epoch": 33.33139365725924, + "grad_norm": 0.15742386877536774, + "learning_rate": 1.683041912194333e-07, + "loss": 0.0128, + "step": 114560 + }, + { + "epoch": 33.33430317137038, + "grad_norm": 0.1359497457742691, + "learning_rate": 1.676872622377812e-07, + "loss": 0.0113, + "step": 114570 + }, + { + "epoch": 33.337212685481525, + "grad_norm": 0.20723038911819458, + "learning_rate": 1.67071459672396e-07, + "loss": 0.0098, + "step": 114580 + }, + { + "epoch": 33.34012219959267, + "grad_norm": 0.12297632545232773, + "learning_rate": 1.6645678357004367e-07, + "loss": 0.0117, + "step": 114590 + }, + { + "epoch": 33.34303171370381, + "grad_norm": 0.0676383301615715, + "learning_rate": 1.6584323397740853e-07, + "loss": 0.0104, + "step": 114600 + }, + { + "epoch": 33.345941227814954, + "grad_norm": 0.17722873389720917, + "learning_rate": 1.6523081094108161e-07, + "loss": 0.011, + "step": 114610 + }, + { + "epoch": 33.3488507419261, + "grad_norm": 0.07686855643987656, + "learning_rate": 1.646195145075724e-07, + "loss": 0.0083, + "step": 114620 + }, + { + "epoch": 33.351760256037245, + "grad_norm": 0.07596100121736526, + "learning_rate": 1.6400934472330708e-07, + "loss": 0.0111, + "step": 114630 + }, + { + "epoch": 33.354669770148384, + "grad_norm": 0.12391330301761627, + "learning_rate": 1.634003016346236e-07, + "loss": 0.0127, + "step": 114640 + }, + { + "epoch": 33.35757928425953, + "grad_norm": 0.10760276764631271, + "learning_rate": 1.627923852877733e-07, + "loss": 0.0096, + "step": 114650 + }, + { + "epoch": 33.360488798370675, + "grad_norm": 0.1168861985206604, + "learning_rate": 1.6218559572892422e-07, + "loss": 0.0107, + "step": 114660 + }, + { + "epoch": 33.36339831248181, + "grad_norm": 0.10089998692274094, + "learning_rate": 1.6157993300415786e-07, + "loss": 0.0098, + "step": 114670 + }, + { + "epoch": 33.36630782659296, + "grad_norm": 0.10503072291612625, + "learning_rate": 1.609753971594724e-07, + "loss": 0.0117, + "step": 114680 + }, + { + "epoch": 33.369217340704104, + "grad_norm": 0.11273769289255142, + "learning_rate": 1.6037198824077447e-07, + "loss": 0.0125, + "step": 114690 + }, + { + "epoch": 33.37212685481525, + "grad_norm": 0.1396297961473465, + "learning_rate": 1.5976970629389076e-07, + "loss": 0.0118, + "step": 114700 + }, + { + "epoch": 33.37503636892639, + "grad_norm": 0.10560968518257141, + "learning_rate": 1.5916855136456298e-07, + "loss": 0.011, + "step": 114710 + }, + { + "epoch": 33.37794588303753, + "grad_norm": 0.09577468037605286, + "learning_rate": 1.5856852349844131e-07, + "loss": 0.0089, + "step": 114720 + }, + { + "epoch": 33.38085539714868, + "grad_norm": 0.14094744622707367, + "learning_rate": 1.5796962274109595e-07, + "loss": 0.0095, + "step": 114730 + }, + { + "epoch": 33.38376491125982, + "grad_norm": 0.12618467211723328, + "learning_rate": 1.5737184913800718e-07, + "loss": 0.011, + "step": 114740 + }, + { + "epoch": 33.38667442537096, + "grad_norm": 0.11381573975086212, + "learning_rate": 1.5677520273457534e-07, + "loss": 0.0117, + "step": 114750 + }, + { + "epoch": 33.38958393948211, + "grad_norm": 0.07933652400970459, + "learning_rate": 1.5617968357610922e-07, + "loss": 0.0108, + "step": 114760 + }, + { + "epoch": 33.39249345359325, + "grad_norm": 0.1643393486738205, + "learning_rate": 1.5558529170783597e-07, + "loss": 0.0123, + "step": 114770 + }, + { + "epoch": 33.39540296770439, + "grad_norm": 0.11940521746873856, + "learning_rate": 1.5499202717489614e-07, + "loss": 0.0117, + "step": 114780 + }, + { + "epoch": 33.39831248181554, + "grad_norm": 0.14546485245227814, + "learning_rate": 1.54399890022342e-07, + "loss": 0.0115, + "step": 114790 + }, + { + "epoch": 33.40122199592668, + "grad_norm": 0.07253575325012207, + "learning_rate": 1.5380888029514428e-07, + "loss": 0.0107, + "step": 114800 + }, + { + "epoch": 33.40413151003782, + "grad_norm": 0.0755431205034256, + "learning_rate": 1.532189980381854e-07, + "loss": 0.011, + "step": 114810 + }, + { + "epoch": 33.40704102414897, + "grad_norm": 0.09480338543653488, + "learning_rate": 1.526302432962645e-07, + "loss": 0.009, + "step": 114820 + }, + { + "epoch": 33.40995053826011, + "grad_norm": 0.08850789815187454, + "learning_rate": 1.5204261611409088e-07, + "loss": 0.0101, + "step": 114830 + }, + { + "epoch": 33.41286005237125, + "grad_norm": 0.08058708161115646, + "learning_rate": 1.5145611653629378e-07, + "loss": 0.0098, + "step": 114840 + }, + { + "epoch": 33.415769566482396, + "grad_norm": 0.08390560001134872, + "learning_rate": 1.5087074460741257e-07, + "loss": 0.009, + "step": 114850 + }, + { + "epoch": 33.41867908059354, + "grad_norm": 0.09424861520528793, + "learning_rate": 1.5028650037190171e-07, + "loss": 0.0093, + "step": 114860 + }, + { + "epoch": 33.42158859470469, + "grad_norm": 0.10516995191574097, + "learning_rate": 1.497033838741324e-07, + "loss": 0.011, + "step": 114870 + }, + { + "epoch": 33.424498108815826, + "grad_norm": 0.1237846314907074, + "learning_rate": 1.4912139515838752e-07, + "loss": 0.011, + "step": 114880 + }, + { + "epoch": 33.42740762292697, + "grad_norm": 0.11270633339881897, + "learning_rate": 1.4854053426886505e-07, + "loss": 0.0092, + "step": 114890 + }, + { + "epoch": 33.43031713703812, + "grad_norm": 0.14799898862838745, + "learning_rate": 1.4796080124967803e-07, + "loss": 0.0118, + "step": 114900 + }, + { + "epoch": 33.433226651149255, + "grad_norm": 0.061912670731544495, + "learning_rate": 1.473821961448546e-07, + "loss": 0.0095, + "step": 114910 + }, + { + "epoch": 33.4361361652604, + "grad_norm": 0.09885569661855698, + "learning_rate": 1.4680471899833292e-07, + "loss": 0.009, + "step": 114920 + }, + { + "epoch": 33.439045679371546, + "grad_norm": 0.10387212038040161, + "learning_rate": 1.4622836985396958e-07, + "loss": 0.0109, + "step": 114930 + }, + { + "epoch": 33.44195519348269, + "grad_norm": 0.10074307024478912, + "learning_rate": 1.4565314875553626e-07, + "loss": 0.0108, + "step": 114940 + }, + { + "epoch": 33.44486470759383, + "grad_norm": 0.11219041794538498, + "learning_rate": 1.4507905574671797e-07, + "loss": 0.0112, + "step": 114950 + }, + { + "epoch": 33.447774221704975, + "grad_norm": 0.08163093775510788, + "learning_rate": 1.445060908711082e-07, + "loss": 0.0093, + "step": 114960 + }, + { + "epoch": 33.45068373581612, + "grad_norm": 0.19116206467151642, + "learning_rate": 1.439342541722255e-07, + "loss": 0.0089, + "step": 114970 + }, + { + "epoch": 33.45359324992726, + "grad_norm": 0.10185519605875015, + "learning_rate": 1.433635456934934e-07, + "loss": 0.0099, + "step": 114980 + }, + { + "epoch": 33.456502764038405, + "grad_norm": 0.12924736738204956, + "learning_rate": 1.4279396547825564e-07, + "loss": 0.012, + "step": 114990 + }, + { + "epoch": 33.45941227814955, + "grad_norm": 0.0821365937590599, + "learning_rate": 1.4222551356976587e-07, + "loss": 0.0112, + "step": 115000 + }, + { + "epoch": 33.462321792260695, + "grad_norm": 0.12972381711006165, + "learning_rate": 1.4165819001119628e-07, + "loss": 0.0093, + "step": 115010 + }, + { + "epoch": 33.465231306371834, + "grad_norm": 0.11479459702968597, + "learning_rate": 1.4109199484563075e-07, + "loss": 0.0095, + "step": 115020 + }, + { + "epoch": 33.46814082048298, + "grad_norm": 0.09311886131763458, + "learning_rate": 1.4052692811606816e-07, + "loss": 0.0111, + "step": 115030 + }, + { + "epoch": 33.471050334594125, + "grad_norm": 0.13909149169921875, + "learning_rate": 1.3996298986542087e-07, + "loss": 0.0095, + "step": 115040 + }, + { + "epoch": 33.47395984870526, + "grad_norm": 0.09061503410339355, + "learning_rate": 1.39400180136518e-07, + "loss": 0.0126, + "step": 115050 + }, + { + "epoch": 33.47686936281641, + "grad_norm": 0.10700366646051407, + "learning_rate": 1.3883849897209865e-07, + "loss": 0.0105, + "step": 115060 + }, + { + "epoch": 33.479778876927554, + "grad_norm": 0.1023109182715416, + "learning_rate": 1.382779464148204e-07, + "loss": 0.0112, + "step": 115070 + }, + { + "epoch": 33.4826883910387, + "grad_norm": 0.20828597247600555, + "learning_rate": 1.3771852250725413e-07, + "loss": 0.0111, + "step": 115080 + }, + { + "epoch": 33.48559790514984, + "grad_norm": 0.08630622178316116, + "learning_rate": 1.371602272918826e-07, + "loss": 0.0098, + "step": 115090 + }, + { + "epoch": 33.48850741926098, + "grad_norm": 0.06674717366695404, + "learning_rate": 1.3660306081110517e-07, + "loss": 0.0114, + "step": 115100 + }, + { + "epoch": 33.49141693337213, + "grad_norm": 0.09136655926704407, + "learning_rate": 1.360470231072364e-07, + "loss": 0.01, + "step": 115110 + }, + { + "epoch": 33.49432644748327, + "grad_norm": 0.1288871169090271, + "learning_rate": 1.354921142225024e-07, + "loss": 0.0096, + "step": 115120 + }, + { + "epoch": 33.49723596159441, + "grad_norm": 0.08691372722387314, + "learning_rate": 1.3493833419904454e-07, + "loss": 0.0094, + "step": 115130 + }, + { + "epoch": 33.50014547570556, + "grad_norm": 0.11058409512042999, + "learning_rate": 1.3438568307892085e-07, + "loss": 0.0112, + "step": 115140 + }, + { + "epoch": 33.503054989816704, + "grad_norm": 0.18543240427970886, + "learning_rate": 1.338341609040994e-07, + "loss": 0.0101, + "step": 115150 + }, + { + "epoch": 33.50596450392784, + "grad_norm": 0.1164708137512207, + "learning_rate": 1.3328376771646334e-07, + "loss": 0.0088, + "step": 115160 + }, + { + "epoch": 33.50887401803899, + "grad_norm": 0.07711146026849747, + "learning_rate": 1.327345035578159e-07, + "loss": 0.0097, + "step": 115170 + }, + { + "epoch": 33.51178353215013, + "grad_norm": 0.11332345753908157, + "learning_rate": 1.3218636846986543e-07, + "loss": 0.0119, + "step": 115180 + }, + { + "epoch": 33.51469304626127, + "grad_norm": 0.20833554863929749, + "learning_rate": 1.316393624942436e-07, + "loss": 0.0117, + "step": 115190 + }, + { + "epoch": 33.51760256037242, + "grad_norm": 0.10343638807535172, + "learning_rate": 1.3109348567248712e-07, + "loss": 0.011, + "step": 115200 + }, + { + "epoch": 33.52051207448356, + "grad_norm": 0.10770288109779358, + "learning_rate": 1.3054873804605627e-07, + "loss": 0.0082, + "step": 115210 + }, + { + "epoch": 33.52342158859471, + "grad_norm": 0.10210409760475159, + "learning_rate": 1.300051196563179e-07, + "loss": 0.0097, + "step": 115220 + }, + { + "epoch": 33.526331102705846, + "grad_norm": 0.07165789604187012, + "learning_rate": 1.294626305445573e-07, + "loss": 0.0099, + "step": 115230 + }, + { + "epoch": 33.52924061681699, + "grad_norm": 0.09223448485136032, + "learning_rate": 1.2892127075197324e-07, + "loss": 0.0101, + "step": 115240 + }, + { + "epoch": 33.53215013092814, + "grad_norm": 0.17054957151412964, + "learning_rate": 1.2838104031967778e-07, + "loss": 0.0126, + "step": 115250 + }, + { + "epoch": 33.535059645039276, + "grad_norm": 0.09729977697134018, + "learning_rate": 1.2784193928869647e-07, + "loss": 0.01, + "step": 115260 + }, + { + "epoch": 33.53796915915042, + "grad_norm": 0.24766743183135986, + "learning_rate": 1.2730396769997487e-07, + "loss": 0.0102, + "step": 115270 + }, + { + "epoch": 33.54087867326157, + "grad_norm": 0.10492187738418579, + "learning_rate": 1.267671255943653e-07, + "loss": 0.0109, + "step": 115280 + }, + { + "epoch": 33.54378818737271, + "grad_norm": 0.12841401994228363, + "learning_rate": 1.2623141301263686e-07, + "loss": 0.0107, + "step": 115290 + }, + { + "epoch": 33.54669770148385, + "grad_norm": 0.06626275926828384, + "learning_rate": 1.256968299954736e-07, + "loss": 0.0074, + "step": 115300 + }, + { + "epoch": 33.549607215594996, + "grad_norm": 0.08722301572561264, + "learning_rate": 1.2516337658347643e-07, + "loss": 0.0098, + "step": 115310 + }, + { + "epoch": 33.55251672970614, + "grad_norm": 0.08827709406614304, + "learning_rate": 1.2463105281715458e-07, + "loss": 0.0093, + "step": 115320 + }, + { + "epoch": 33.55542624381728, + "grad_norm": 0.09700597077608109, + "learning_rate": 1.240998587369341e-07, + "loss": 0.0106, + "step": 115330 + }, + { + "epoch": 33.558335757928425, + "grad_norm": 0.08102884888648987, + "learning_rate": 1.2356979438315764e-07, + "loss": 0.0093, + "step": 115340 + }, + { + "epoch": 33.56124527203957, + "grad_norm": 0.09546265006065369, + "learning_rate": 1.230408597960797e-07, + "loss": 0.0108, + "step": 115350 + }, + { + "epoch": 33.56415478615071, + "grad_norm": 0.0673157274723053, + "learning_rate": 1.2251305501586818e-07, + "loss": 0.0121, + "step": 115360 + }, + { + "epoch": 33.567064300261855, + "grad_norm": 0.11667505651712418, + "learning_rate": 1.2198638008260765e-07, + "loss": 0.0108, + "step": 115370 + }, + { + "epoch": 33.569973814373, + "grad_norm": 0.14977186918258667, + "learning_rate": 1.2146083503629612e-07, + "loss": 0.0115, + "step": 115380 + }, + { + "epoch": 33.572883328484146, + "grad_norm": 0.07893355190753937, + "learning_rate": 1.20936419916845e-07, + "loss": 0.0104, + "step": 115390 + }, + { + "epoch": 33.575792842595284, + "grad_norm": 0.13183772563934326, + "learning_rate": 1.2041313476407744e-07, + "loss": 0.0105, + "step": 115400 + }, + { + "epoch": 33.57870235670643, + "grad_norm": 0.15581096708774567, + "learning_rate": 1.1989097961773666e-07, + "loss": 0.0109, + "step": 115410 + }, + { + "epoch": 33.581611870817575, + "grad_norm": 0.1550799459218979, + "learning_rate": 1.193699545174759e-07, + "loss": 0.0111, + "step": 115420 + }, + { + "epoch": 33.58452138492871, + "grad_norm": 0.08831837773323059, + "learning_rate": 1.1885005950286354e-07, + "loss": 0.011, + "step": 115430 + }, + { + "epoch": 33.58743089903986, + "grad_norm": 0.12038827687501907, + "learning_rate": 1.1833129461338133e-07, + "loss": 0.0104, + "step": 115440 + }, + { + "epoch": 33.590340413151004, + "grad_norm": 0.2220965474843979, + "learning_rate": 1.1781365988842774e-07, + "loss": 0.0096, + "step": 115450 + }, + { + "epoch": 33.59324992726215, + "grad_norm": 0.12044229358434677, + "learning_rate": 1.1729715536731134e-07, + "loss": 0.0113, + "step": 115460 + }, + { + "epoch": 33.59615944137329, + "grad_norm": 0.19002002477645874, + "learning_rate": 1.1678178108926075e-07, + "loss": 0.0114, + "step": 115470 + }, + { + "epoch": 33.599068955484434, + "grad_norm": 0.06616521626710892, + "learning_rate": 1.1626753709341132e-07, + "loss": 0.009, + "step": 115480 + }, + { + "epoch": 33.60197846959558, + "grad_norm": 0.11439889669418335, + "learning_rate": 1.1575442341882014e-07, + "loss": 0.0093, + "step": 115490 + }, + { + "epoch": 33.60488798370672, + "grad_norm": 0.158896803855896, + "learning_rate": 1.1524244010445106e-07, + "loss": 0.0098, + "step": 115500 + }, + { + "epoch": 33.60779749781786, + "grad_norm": 0.10217086970806122, + "learning_rate": 1.1473158718918963e-07, + "loss": 0.0115, + "step": 115510 + }, + { + "epoch": 33.61070701192901, + "grad_norm": 0.10855524241924286, + "learning_rate": 1.1422186471182816e-07, + "loss": 0.0122, + "step": 115520 + }, + { + "epoch": 33.613616526040154, + "grad_norm": 0.128092423081398, + "learning_rate": 1.1371327271107901e-07, + "loss": 0.0116, + "step": 115530 + }, + { + "epoch": 33.61652604015129, + "grad_norm": 0.08798848092556, + "learning_rate": 1.1320581122556628e-07, + "loss": 0.0095, + "step": 115540 + }, + { + "epoch": 33.61943555426244, + "grad_norm": 0.14515765011310577, + "learning_rate": 1.1269948029382915e-07, + "loss": 0.0107, + "step": 115550 + }, + { + "epoch": 33.62234506837358, + "grad_norm": 0.10591159760951996, + "learning_rate": 1.1219427995431685e-07, + "loss": 0.0102, + "step": 115560 + }, + { + "epoch": 33.62525458248472, + "grad_norm": 0.15864084661006927, + "learning_rate": 1.1169021024539872e-07, + "loss": 0.0101, + "step": 115570 + }, + { + "epoch": 33.62816409659587, + "grad_norm": 0.07464896887540817, + "learning_rate": 1.111872712053541e-07, + "loss": 0.0109, + "step": 115580 + }, + { + "epoch": 33.63107361070701, + "grad_norm": 0.08895702660083771, + "learning_rate": 1.1068546287237913e-07, + "loss": 0.0101, + "step": 115590 + }, + { + "epoch": 33.63398312481816, + "grad_norm": 0.15505990386009216, + "learning_rate": 1.1018478528458332e-07, + "loss": 0.0122, + "step": 115600 + }, + { + "epoch": 33.6368926389293, + "grad_norm": 0.14215737581253052, + "learning_rate": 1.096852384799879e-07, + "loss": 0.01, + "step": 115610 + }, + { + "epoch": 33.63980215304044, + "grad_norm": 0.08754459768533707, + "learning_rate": 1.0918682249653255e-07, + "loss": 0.0095, + "step": 115620 + }, + { + "epoch": 33.64271166715159, + "grad_norm": 0.08705062419176102, + "learning_rate": 1.0868953737206533e-07, + "loss": 0.0097, + "step": 115630 + }, + { + "epoch": 33.645621181262726, + "grad_norm": 0.1044706329703331, + "learning_rate": 1.0819338314435267e-07, + "loss": 0.0102, + "step": 115640 + }, + { + "epoch": 33.64853069537387, + "grad_norm": 0.09619296342134476, + "learning_rate": 1.0769835985107779e-07, + "loss": 0.0105, + "step": 115650 + }, + { + "epoch": 33.65144020948502, + "grad_norm": 0.09079606086015701, + "learning_rate": 1.072044675298306e-07, + "loss": 0.0105, + "step": 115660 + }, + { + "epoch": 33.65434972359616, + "grad_norm": 0.08917015790939331, + "learning_rate": 1.0671170621811943e-07, + "loss": 0.0096, + "step": 115670 + }, + { + "epoch": 33.6572592377073, + "grad_norm": 0.10294067859649658, + "learning_rate": 1.0622007595336936e-07, + "loss": 0.0118, + "step": 115680 + }, + { + "epoch": 33.660168751818446, + "grad_norm": 0.08469512313604355, + "learning_rate": 1.0572957677291217e-07, + "loss": 0.0102, + "step": 115690 + }, + { + "epoch": 33.66307826592959, + "grad_norm": 0.0809379369020462, + "learning_rate": 1.0524020871400141e-07, + "loss": 0.0079, + "step": 115700 + }, + { + "epoch": 33.66598778004073, + "grad_norm": 0.1469736546278, + "learning_rate": 1.0475197181379903e-07, + "loss": 0.0111, + "step": 115710 + }, + { + "epoch": 33.668897294151876, + "grad_norm": 0.09385460615158081, + "learning_rate": 1.0426486610938368e-07, + "loss": 0.0097, + "step": 115720 + }, + { + "epoch": 33.67180680826302, + "grad_norm": 0.12486802786588669, + "learning_rate": 1.0377889163774911e-07, + "loss": 0.0104, + "step": 115730 + }, + { + "epoch": 33.67471632237417, + "grad_norm": 0.09400205314159393, + "learning_rate": 1.0329404843580248e-07, + "loss": 0.011, + "step": 115740 + }, + { + "epoch": 33.677625836485305, + "grad_norm": 0.15378637611865997, + "learning_rate": 1.0281033654036098e-07, + "loss": 0.0107, + "step": 115750 + }, + { + "epoch": 33.68053535059645, + "grad_norm": 0.052655190229415894, + "learning_rate": 1.0232775598816358e-07, + "loss": 0.0105, + "step": 115760 + }, + { + "epoch": 33.683444864707596, + "grad_norm": 0.1564498394727707, + "learning_rate": 1.0184630681585427e-07, + "loss": 0.0102, + "step": 115770 + }, + { + "epoch": 33.686354378818734, + "grad_norm": 0.09940314292907715, + "learning_rate": 1.013659890600005e-07, + "loss": 0.0097, + "step": 115780 + }, + { + "epoch": 33.68926389292988, + "grad_norm": 0.0868314728140831, + "learning_rate": 1.0088680275707807e-07, + "loss": 0.0111, + "step": 115790 + }, + { + "epoch": 33.692173407041025, + "grad_norm": 0.2319600135087967, + "learning_rate": 1.004087479434762e-07, + "loss": 0.0139, + "step": 115800 + }, + { + "epoch": 33.69508292115217, + "grad_norm": 0.08892180770635605, + "learning_rate": 9.993182465550254e-08, + "loss": 0.0095, + "step": 115810 + }, + { + "epoch": 33.69799243526331, + "grad_norm": 0.12556962668895721, + "learning_rate": 9.945603292937311e-08, + "loss": 0.0098, + "step": 115820 + }, + { + "epoch": 33.700901949374455, + "grad_norm": 0.09136494994163513, + "learning_rate": 9.8981372801224e-08, + "loss": 0.0096, + "step": 115830 + }, + { + "epoch": 33.7038114634856, + "grad_norm": 0.14384965598583221, + "learning_rate": 9.850784430710135e-08, + "loss": 0.0103, + "step": 115840 + }, + { + "epoch": 33.70672097759674, + "grad_norm": 0.1305461972951889, + "learning_rate": 9.803544748296644e-08, + "loss": 0.0116, + "step": 115850 + }, + { + "epoch": 33.709630491707884, + "grad_norm": 0.13518047332763672, + "learning_rate": 9.756418236469555e-08, + "loss": 0.0091, + "step": 115860 + }, + { + "epoch": 33.71254000581903, + "grad_norm": 0.11526558548212051, + "learning_rate": 9.709404898807673e-08, + "loss": 0.0123, + "step": 115870 + }, + { + "epoch": 33.715449519930175, + "grad_norm": 0.07993466407060623, + "learning_rate": 9.662504738881473e-08, + "loss": 0.0107, + "step": 115880 + }, + { + "epoch": 33.71835903404131, + "grad_norm": 0.09202495217323303, + "learning_rate": 9.615717760252607e-08, + "loss": 0.0123, + "step": 115890 + }, + { + "epoch": 33.72126854815246, + "grad_norm": 0.08900069445371628, + "learning_rate": 9.569043966474233e-08, + "loss": 0.0115, + "step": 115900 + }, + { + "epoch": 33.724178062263604, + "grad_norm": 0.13634629547595978, + "learning_rate": 9.522483361091183e-08, + "loss": 0.0098, + "step": 115910 + }, + { + "epoch": 33.72708757637474, + "grad_norm": 0.1077101081609726, + "learning_rate": 9.476035947639128e-08, + "loss": 0.0103, + "step": 115920 + }, + { + "epoch": 33.72999709048589, + "grad_norm": 0.12857982516288757, + "learning_rate": 9.429701729645413e-08, + "loss": 0.0083, + "step": 115930 + }, + { + "epoch": 33.732906604597034, + "grad_norm": 0.10815288871526718, + "learning_rate": 9.383480710629056e-08, + "loss": 0.0087, + "step": 115940 + }, + { + "epoch": 33.73581611870818, + "grad_norm": 0.0890195369720459, + "learning_rate": 9.337372894100248e-08, + "loss": 0.0096, + "step": 115950 + }, + { + "epoch": 33.73872563281932, + "grad_norm": 0.0802723690867424, + "learning_rate": 9.291378283560358e-08, + "loss": 0.0097, + "step": 115960 + }, + { + "epoch": 33.74163514693046, + "grad_norm": 0.1255294382572174, + "learning_rate": 9.245496882502258e-08, + "loss": 0.0102, + "step": 115970 + }, + { + "epoch": 33.74454466104161, + "grad_norm": 0.10137420892715454, + "learning_rate": 9.199728694410825e-08, + "loss": 0.0115, + "step": 115980 + }, + { + "epoch": 33.74745417515275, + "grad_norm": 0.11305143684148788, + "learning_rate": 9.154073722761613e-08, + "loss": 0.0114, + "step": 115990 + }, + { + "epoch": 33.75036368926389, + "grad_norm": 0.09899803251028061, + "learning_rate": 9.108531971021516e-08, + "loss": 0.0092, + "step": 116000 + }, + { + "epoch": 33.75327320337504, + "grad_norm": 0.1314404457807541, + "learning_rate": 9.063103442649434e-08, + "loss": 0.0115, + "step": 116010 + }, + { + "epoch": 33.75618271748618, + "grad_norm": 0.1122356727719307, + "learning_rate": 9.01778814109544e-08, + "loss": 0.0104, + "step": 116020 + }, + { + "epoch": 33.75909223159732, + "grad_norm": 0.11307726800441742, + "learning_rate": 8.972586069800948e-08, + "loss": 0.0089, + "step": 116030 + }, + { + "epoch": 33.76200174570847, + "grad_norm": 0.148495152592659, + "learning_rate": 8.927497232198379e-08, + "loss": 0.0109, + "step": 116040 + }, + { + "epoch": 33.76491125981961, + "grad_norm": 0.12027471512556076, + "learning_rate": 8.88252163171216e-08, + "loss": 0.0098, + "step": 116050 + }, + { + "epoch": 33.76782077393075, + "grad_norm": 0.2736227810382843, + "learning_rate": 8.837659271758058e-08, + "loss": 0.0098, + "step": 116060 + }, + { + "epoch": 33.7707302880419, + "grad_norm": 0.11919072270393372, + "learning_rate": 8.79291015574285e-08, + "loss": 0.0112, + "step": 116070 + }, + { + "epoch": 33.77363980215304, + "grad_norm": 0.09010079503059387, + "learning_rate": 8.748274287064817e-08, + "loss": 0.0116, + "step": 116080 + }, + { + "epoch": 33.77654931626418, + "grad_norm": 0.1027393788099289, + "learning_rate": 8.703751669114246e-08, + "loss": 0.0094, + "step": 116090 + }, + { + "epoch": 33.779458830375326, + "grad_norm": 0.10634832084178925, + "learning_rate": 8.659342305271933e-08, + "loss": 0.0102, + "step": 116100 + }, + { + "epoch": 33.78236834448647, + "grad_norm": 0.11235425621271133, + "learning_rate": 8.615046198910514e-08, + "loss": 0.0102, + "step": 116110 + }, + { + "epoch": 33.78527785859762, + "grad_norm": 0.09590958803892136, + "learning_rate": 8.570863353394132e-08, + "loss": 0.0097, + "step": 116120 + }, + { + "epoch": 33.788187372708755, + "grad_norm": 0.10426904261112213, + "learning_rate": 8.526793772078101e-08, + "loss": 0.0099, + "step": 116130 + }, + { + "epoch": 33.7910968868199, + "grad_norm": 0.11052341759204865, + "learning_rate": 8.482837458309079e-08, + "loss": 0.0108, + "step": 116140 + }, + { + "epoch": 33.794006400931046, + "grad_norm": 0.10797492414712906, + "learning_rate": 8.43899441542556e-08, + "loss": 0.0108, + "step": 116150 + }, + { + "epoch": 33.796915915042185, + "grad_norm": 0.12319188565015793, + "learning_rate": 8.395264646756884e-08, + "loss": 0.0101, + "step": 116160 + }, + { + "epoch": 33.79982542915333, + "grad_norm": 0.11790390312671661, + "learning_rate": 8.351648155624225e-08, + "loss": 0.0113, + "step": 116170 + }, + { + "epoch": 33.802734943264475, + "grad_norm": 0.10497536510229111, + "learning_rate": 8.308144945339768e-08, + "loss": 0.0114, + "step": 116180 + }, + { + "epoch": 33.80564445737562, + "grad_norm": 0.10254865884780884, + "learning_rate": 8.26475501920737e-08, + "loss": 0.0114, + "step": 116190 + }, + { + "epoch": 33.80855397148676, + "grad_norm": 0.11398756504058838, + "learning_rate": 8.22147838052223e-08, + "loss": 0.0111, + "step": 116200 + }, + { + "epoch": 33.811463485597905, + "grad_norm": 0.060443777590990067, + "learning_rate": 8.178315032571049e-08, + "loss": 0.0119, + "step": 116210 + }, + { + "epoch": 33.81437299970905, + "grad_norm": 0.13265180587768555, + "learning_rate": 8.135264978631707e-08, + "loss": 0.0105, + "step": 116220 + }, + { + "epoch": 33.81728251382019, + "grad_norm": 0.12997151911258698, + "learning_rate": 8.092328221973256e-08, + "loss": 0.0108, + "step": 116230 + }, + { + "epoch": 33.820192027931334, + "grad_norm": 0.10820750147104263, + "learning_rate": 8.04950476585692e-08, + "loss": 0.0089, + "step": 116240 + }, + { + "epoch": 33.82310154204248, + "grad_norm": 0.08593837916851044, + "learning_rate": 8.006794613534762e-08, + "loss": 0.0091, + "step": 116250 + }, + { + "epoch": 33.826011056153625, + "grad_norm": 0.1591130793094635, + "learning_rate": 7.964197768250192e-08, + "loss": 0.009, + "step": 116260 + }, + { + "epoch": 33.82892057026476, + "grad_norm": 0.07031402736902237, + "learning_rate": 7.92171423323812e-08, + "loss": 0.0087, + "step": 116270 + }, + { + "epoch": 33.83183008437591, + "grad_norm": 0.07973747700452805, + "learning_rate": 7.879344011725132e-08, + "loss": 0.0102, + "step": 116280 + }, + { + "epoch": 33.834739598487054, + "grad_norm": 0.07641232013702393, + "learning_rate": 7.83708710692882e-08, + "loss": 0.0085, + "step": 116290 + }, + { + "epoch": 33.83764911259819, + "grad_norm": 0.12264532595872879, + "learning_rate": 7.794943522058285e-08, + "loss": 0.0095, + "step": 116300 + }, + { + "epoch": 33.84055862670934, + "grad_norm": 0.09446091949939728, + "learning_rate": 7.752913260313965e-08, + "loss": 0.0097, + "step": 116310 + }, + { + "epoch": 33.843468140820484, + "grad_norm": 0.06678557395935059, + "learning_rate": 7.710996324888142e-08, + "loss": 0.0094, + "step": 116320 + }, + { + "epoch": 33.84637765493163, + "grad_norm": 0.07549712806940079, + "learning_rate": 7.669192718963769e-08, + "loss": 0.0096, + "step": 116330 + }, + { + "epoch": 33.84928716904277, + "grad_norm": 0.14557233452796936, + "learning_rate": 7.62750244571564e-08, + "loss": 0.0106, + "step": 116340 + }, + { + "epoch": 33.85219668315391, + "grad_norm": 0.10042840242385864, + "learning_rate": 7.585925508309887e-08, + "loss": 0.0106, + "step": 116350 + }, + { + "epoch": 33.85510619726506, + "grad_norm": 0.15194033086299896, + "learning_rate": 7.544461909904154e-08, + "loss": 0.0101, + "step": 116360 + }, + { + "epoch": 33.8580157113762, + "grad_norm": 0.09395325183868408, + "learning_rate": 7.503111653647087e-08, + "loss": 0.0098, + "step": 116370 + }, + { + "epoch": 33.86092522548734, + "grad_norm": 0.1583215594291687, + "learning_rate": 7.46187474267901e-08, + "loss": 0.0117, + "step": 116380 + }, + { + "epoch": 33.86383473959849, + "grad_norm": 0.06531597673892975, + "learning_rate": 7.420751180131913e-08, + "loss": 0.0103, + "step": 116390 + }, + { + "epoch": 33.86674425370963, + "grad_norm": 0.0869482159614563, + "learning_rate": 7.379740969128468e-08, + "loss": 0.0089, + "step": 116400 + }, + { + "epoch": 33.86965376782077, + "grad_norm": 0.14112251996994019, + "learning_rate": 7.338844112783182e-08, + "loss": 0.0105, + "step": 116410 + }, + { + "epoch": 33.87256328193192, + "grad_norm": 0.0909552350640297, + "learning_rate": 7.298060614201907e-08, + "loss": 0.0091, + "step": 116420 + }, + { + "epoch": 33.87547279604306, + "grad_norm": 0.07424934953451157, + "learning_rate": 7.257390476482162e-08, + "loss": 0.0105, + "step": 116430 + }, + { + "epoch": 33.8783823101542, + "grad_norm": 0.1659509688615799, + "learning_rate": 7.216833702712145e-08, + "loss": 0.0118, + "step": 116440 + }, + { + "epoch": 33.88129182426535, + "grad_norm": 0.14968298375606537, + "learning_rate": 7.176390295972223e-08, + "loss": 0.0113, + "step": 116450 + }, + { + "epoch": 33.88420133837649, + "grad_norm": 0.1086306944489479, + "learning_rate": 7.136060259333776e-08, + "loss": 0.0102, + "step": 116460 + }, + { + "epoch": 33.88711085248764, + "grad_norm": 0.08870009332895279, + "learning_rate": 7.095843595859352e-08, + "loss": 0.0119, + "step": 116470 + }, + { + "epoch": 33.890020366598776, + "grad_norm": 0.11762186139822006, + "learning_rate": 7.055740308603342e-08, + "loss": 0.0109, + "step": 116480 + }, + { + "epoch": 33.89292988070992, + "grad_norm": 0.14728033542633057, + "learning_rate": 7.015750400611309e-08, + "loss": 0.0091, + "step": 116490 + }, + { + "epoch": 33.89583939482107, + "grad_norm": 0.11433687806129456, + "learning_rate": 6.975873874920158e-08, + "loss": 0.0124, + "step": 116500 + }, + { + "epoch": 33.898748908932205, + "grad_norm": 0.09055769443511963, + "learning_rate": 6.936110734558298e-08, + "loss": 0.0084, + "step": 116510 + }, + { + "epoch": 33.90165842304335, + "grad_norm": 0.11175937950611115, + "learning_rate": 6.896460982545483e-08, + "loss": 0.0102, + "step": 116520 + }, + { + "epoch": 33.904567937154496, + "grad_norm": 0.06953081488609314, + "learning_rate": 6.856924621892802e-08, + "loss": 0.0109, + "step": 116530 + }, + { + "epoch": 33.90747745126564, + "grad_norm": 0.1447017937898636, + "learning_rate": 6.817501655602687e-08, + "loss": 0.01, + "step": 116540 + }, + { + "epoch": 33.91038696537678, + "grad_norm": 0.11325273662805557, + "learning_rate": 6.77819208666941e-08, + "loss": 0.0093, + "step": 116550 + }, + { + "epoch": 33.913296479487926, + "grad_norm": 0.10696079581975937, + "learning_rate": 6.738995918077916e-08, + "loss": 0.0104, + "step": 116560 + }, + { + "epoch": 33.91620599359907, + "grad_norm": 0.1093638688325882, + "learning_rate": 6.699913152804992e-08, + "loss": 0.0125, + "step": 116570 + }, + { + "epoch": 33.91911550771021, + "grad_norm": 0.13859888911247253, + "learning_rate": 6.660943793818764e-08, + "loss": 0.009, + "step": 116580 + }, + { + "epoch": 33.922025021821355, + "grad_norm": 0.11925400048494339, + "learning_rate": 6.622087844078862e-08, + "loss": 0.0106, + "step": 116590 + }, + { + "epoch": 33.9249345359325, + "grad_norm": 0.08783354610204697, + "learning_rate": 6.58334530653576e-08, + "loss": 0.0102, + "step": 116600 + }, + { + "epoch": 33.927844050043646, + "grad_norm": 0.07420484721660614, + "learning_rate": 6.544716184131772e-08, + "loss": 0.0073, + "step": 116610 + }, + { + "epoch": 33.930753564154784, + "grad_norm": 0.17166242003440857, + "learning_rate": 6.506200479800883e-08, + "loss": 0.0115, + "step": 116620 + }, + { + "epoch": 33.93366307826593, + "grad_norm": 0.08142068982124329, + "learning_rate": 6.467798196467755e-08, + "loss": 0.0095, + "step": 116630 + }, + { + "epoch": 33.936572592377075, + "grad_norm": 0.10998876392841339, + "learning_rate": 6.429509337048889e-08, + "loss": 0.0091, + "step": 116640 + }, + { + "epoch": 33.939482106488214, + "grad_norm": 0.07880889624357224, + "learning_rate": 6.391333904452124e-08, + "loss": 0.0091, + "step": 116650 + }, + { + "epoch": 33.94239162059936, + "grad_norm": 0.12332505732774734, + "learning_rate": 6.353271901576641e-08, + "loss": 0.0103, + "step": 116660 + }, + { + "epoch": 33.945301134710505, + "grad_norm": 0.14866307377815247, + "learning_rate": 6.315323331312794e-08, + "loss": 0.0109, + "step": 116670 + }, + { + "epoch": 33.94821064882165, + "grad_norm": 0.07773975282907486, + "learning_rate": 6.27748819654278e-08, + "loss": 0.0086, + "step": 116680 + }, + { + "epoch": 33.95112016293279, + "grad_norm": 0.08883264660835266, + "learning_rate": 6.239766500139799e-08, + "loss": 0.0112, + "step": 116690 + }, + { + "epoch": 33.954029677043934, + "grad_norm": 0.09286633878946304, + "learning_rate": 6.202158244968559e-08, + "loss": 0.0131, + "step": 116700 + }, + { + "epoch": 33.95693919115508, + "grad_norm": 0.09783020615577698, + "learning_rate": 6.164663433885276e-08, + "loss": 0.0111, + "step": 116710 + }, + { + "epoch": 33.95984870526622, + "grad_norm": 0.10943752527236938, + "learning_rate": 6.127282069737172e-08, + "loss": 0.0094, + "step": 116720 + }, + { + "epoch": 33.96275821937736, + "grad_norm": 0.1339055895805359, + "learning_rate": 6.090014155363477e-08, + "loss": 0.0104, + "step": 116730 + }, + { + "epoch": 33.96566773348851, + "grad_norm": 0.07814466208219528, + "learning_rate": 6.05285969359426e-08, + "loss": 0.0092, + "step": 116740 + }, + { + "epoch": 33.968577247599654, + "grad_norm": 0.09179828315973282, + "learning_rate": 6.015818687251096e-08, + "loss": 0.011, + "step": 116750 + }, + { + "epoch": 33.97148676171079, + "grad_norm": 0.07684362679719925, + "learning_rate": 5.978891139147069e-08, + "loss": 0.0089, + "step": 116760 + }, + { + "epoch": 33.97439627582194, + "grad_norm": 0.07612022012472153, + "learning_rate": 5.942077052086603e-08, + "loss": 0.0096, + "step": 116770 + }, + { + "epoch": 33.977305789933084, + "grad_norm": 0.08214996755123138, + "learning_rate": 5.905376428865461e-08, + "loss": 0.0085, + "step": 116780 + }, + { + "epoch": 33.98021530404422, + "grad_norm": 0.100008524954319, + "learning_rate": 5.868789272270747e-08, + "loss": 0.0101, + "step": 116790 + }, + { + "epoch": 33.98312481815537, + "grad_norm": 0.10959991067647934, + "learning_rate": 5.832315585081072e-08, + "loss": 0.0107, + "step": 116800 + }, + { + "epoch": 33.98603433226651, + "grad_norm": 0.10395703464746475, + "learning_rate": 5.7959553700665526e-08, + "loss": 0.0106, + "step": 116810 + }, + { + "epoch": 33.98894384637765, + "grad_norm": 0.07858702540397644, + "learning_rate": 5.759708629988148e-08, + "loss": 0.0092, + "step": 116820 + }, + { + "epoch": 33.9918533604888, + "grad_norm": 0.11881201714277267, + "learning_rate": 5.723575367598821e-08, + "loss": 0.009, + "step": 116830 + }, + { + "epoch": 33.99476287459994, + "grad_norm": 0.09513601660728455, + "learning_rate": 5.687555585642546e-08, + "loss": 0.0092, + "step": 116840 + }, + { + "epoch": 33.99767238871109, + "grad_norm": 0.11122529953718185, + "learning_rate": 5.651649286854799e-08, + "loss": 0.011, + "step": 116850 + }, + { + "epoch": 34.000581902822226, + "grad_norm": 0.10640662163496017, + "learning_rate": 5.615856473962566e-08, + "loss": 0.0102, + "step": 116860 + }, + { + "epoch": 34.00349141693337, + "grad_norm": 0.17757506668567657, + "learning_rate": 5.580177149683841e-08, + "loss": 0.0108, + "step": 116870 + }, + { + "epoch": 34.00640093104452, + "grad_norm": 0.1049092561006546, + "learning_rate": 5.544611316728454e-08, + "loss": 0.0101, + "step": 116880 + }, + { + "epoch": 34.009310445155656, + "grad_norm": 0.09369147568941116, + "learning_rate": 5.5091589777972464e-08, + "loss": 0.0109, + "step": 116890 + }, + { + "epoch": 34.0122199592668, + "grad_norm": 0.11882789433002472, + "learning_rate": 5.4738201355825635e-08, + "loss": 0.0098, + "step": 116900 + }, + { + "epoch": 34.01512947337795, + "grad_norm": 0.06346320360898972, + "learning_rate": 5.438594792768259e-08, + "loss": 0.0095, + "step": 116910 + }, + { + "epoch": 34.01803898748909, + "grad_norm": 0.08617907762527466, + "learning_rate": 5.403482952029526e-08, + "loss": 0.0122, + "step": 116920 + }, + { + "epoch": 34.02094850160023, + "grad_norm": 0.08736756443977356, + "learning_rate": 5.3684846160327314e-08, + "loss": 0.0088, + "step": 116930 + }, + { + "epoch": 34.023858015711376, + "grad_norm": 0.09749122709035873, + "learning_rate": 5.33359978743575e-08, + "loss": 0.0098, + "step": 116940 + }, + { + "epoch": 34.02676752982252, + "grad_norm": 0.10177727788686752, + "learning_rate": 5.2988284688879616e-08, + "loss": 0.0109, + "step": 116950 + }, + { + "epoch": 34.02967704393366, + "grad_norm": 0.10103092342615128, + "learning_rate": 5.2641706630299214e-08, + "loss": 0.0093, + "step": 116960 + }, + { + "epoch": 34.032586558044805, + "grad_norm": 0.1248183324933052, + "learning_rate": 5.229626372493856e-08, + "loss": 0.0096, + "step": 116970 + }, + { + "epoch": 34.03549607215595, + "grad_norm": 0.08026779443025589, + "learning_rate": 5.1951955999028354e-08, + "loss": 0.0106, + "step": 116980 + }, + { + "epoch": 34.038405586267096, + "grad_norm": 0.12426438182592392, + "learning_rate": 5.1608783478721e-08, + "loss": 0.0104, + "step": 116990 + }, + { + "epoch": 34.041315100378235, + "grad_norm": 0.19637510180473328, + "learning_rate": 5.126674619007565e-08, + "loss": 0.0095, + "step": 117000 + }, + { + "epoch": 34.04422461448938, + "grad_norm": 0.08662225306034088, + "learning_rate": 5.092584415906654e-08, + "loss": 0.0102, + "step": 117010 + }, + { + "epoch": 34.047134128600526, + "grad_norm": 0.08472830802202225, + "learning_rate": 5.058607741158461e-08, + "loss": 0.0122, + "step": 117020 + }, + { + "epoch": 34.050043642711664, + "grad_norm": 0.08777248114347458, + "learning_rate": 5.0247445973432557e-08, + "loss": 0.0107, + "step": 117030 + }, + { + "epoch": 34.05295315682281, + "grad_norm": 0.10526551306247711, + "learning_rate": 4.990994987032649e-08, + "loss": 0.0121, + "step": 117040 + }, + { + "epoch": 34.055862670933955, + "grad_norm": 0.14186517894268036, + "learning_rate": 4.9573589127899225e-08, + "loss": 0.0102, + "step": 117050 + }, + { + "epoch": 34.0587721850451, + "grad_norm": 0.11162682622671127, + "learning_rate": 4.9238363771691994e-08, + "loss": 0.011, + "step": 117060 + }, + { + "epoch": 34.06168169915624, + "grad_norm": 0.09783017635345459, + "learning_rate": 4.890427382716611e-08, + "loss": 0.0085, + "step": 117070 + }, + { + "epoch": 34.064591213267384, + "grad_norm": 0.11255001276731491, + "learning_rate": 4.8571319319691276e-08, + "loss": 0.0121, + "step": 117080 + }, + { + "epoch": 34.06750072737853, + "grad_norm": 0.09026402235031128, + "learning_rate": 4.8239500274553924e-08, + "loss": 0.0083, + "step": 117090 + }, + { + "epoch": 34.07041024148967, + "grad_norm": 0.14565962553024292, + "learning_rate": 4.790881671695391e-08, + "loss": 0.0091, + "step": 117100 + }, + { + "epoch": 34.073319755600814, + "grad_norm": 0.11680352687835693, + "learning_rate": 4.75792686720028e-08, + "loss": 0.0094, + "step": 117110 + }, + { + "epoch": 34.07622926971196, + "grad_norm": 0.09266991913318634, + "learning_rate": 4.725085616473057e-08, + "loss": 0.011, + "step": 117120 + }, + { + "epoch": 34.079138783823105, + "grad_norm": 0.075086310505867, + "learning_rate": 4.6923579220075616e-08, + "loss": 0.0104, + "step": 117130 + }, + { + "epoch": 34.08204829793424, + "grad_norm": 0.16648785769939423, + "learning_rate": 4.659743786289306e-08, + "loss": 0.0131, + "step": 117140 + }, + { + "epoch": 34.08495781204539, + "grad_norm": 0.087431401014328, + "learning_rate": 4.627243211794974e-08, + "loss": 0.0105, + "step": 117150 + }, + { + "epoch": 34.087867326156534, + "grad_norm": 0.08936674892902374, + "learning_rate": 4.5948562009930915e-08, + "loss": 0.011, + "step": 117160 + }, + { + "epoch": 34.09077684026767, + "grad_norm": 0.11593475937843323, + "learning_rate": 4.562582756343026e-08, + "loss": 0.0087, + "step": 117170 + }, + { + "epoch": 34.09368635437882, + "grad_norm": 0.12605254352092743, + "learning_rate": 4.530422880295815e-08, + "loss": 0.0107, + "step": 117180 + }, + { + "epoch": 34.09659586848996, + "grad_norm": 0.25290194153785706, + "learning_rate": 4.498376575293672e-08, + "loss": 0.0115, + "step": 117190 + }, + { + "epoch": 34.09950538260111, + "grad_norm": 0.10813874751329422, + "learning_rate": 4.466443843770485e-08, + "loss": 0.0101, + "step": 117200 + }, + { + "epoch": 34.10241489671225, + "grad_norm": 0.07814999669790268, + "learning_rate": 4.434624688150979e-08, + "loss": 0.009, + "step": 117210 + }, + { + "epoch": 34.10532441082339, + "grad_norm": 0.1357916295528412, + "learning_rate": 4.402919110852055e-08, + "loss": 0.0095, + "step": 117220 + }, + { + "epoch": 34.10823392493454, + "grad_norm": 0.12209151685237885, + "learning_rate": 4.371327114281454e-08, + "loss": 0.0108, + "step": 117230 + }, + { + "epoch": 34.11114343904568, + "grad_norm": 0.12207381427288055, + "learning_rate": 4.33984870083809e-08, + "loss": 0.0081, + "step": 117240 + }, + { + "epoch": 34.11405295315682, + "grad_norm": 0.10789269208908081, + "learning_rate": 4.3084838729127165e-08, + "loss": 0.0097, + "step": 117250 + }, + { + "epoch": 34.11696246726797, + "grad_norm": 0.11414363980293274, + "learning_rate": 4.2772326328872624e-08, + "loss": 0.0093, + "step": 117260 + }, + { + "epoch": 34.11987198137911, + "grad_norm": 0.06184956803917885, + "learning_rate": 4.246094983135163e-08, + "loss": 0.0108, + "step": 117270 + }, + { + "epoch": 34.12278149549025, + "grad_norm": 0.09838224202394485, + "learning_rate": 4.215070926021025e-08, + "loss": 0.0105, + "step": 117280 + }, + { + "epoch": 34.1256910096014, + "grad_norm": 0.08186540752649307, + "learning_rate": 4.1841604639009654e-08, + "loss": 0.0094, + "step": 117290 + }, + { + "epoch": 34.12860052371254, + "grad_norm": 0.0737772285938263, + "learning_rate": 4.153363599122273e-08, + "loss": 0.0127, + "step": 117300 + }, + { + "epoch": 34.13151003782368, + "grad_norm": 0.1706942468881607, + "learning_rate": 4.122680334024076e-08, + "loss": 0.0128, + "step": 117310 + }, + { + "epoch": 34.134419551934826, + "grad_norm": 0.10951732099056244, + "learning_rate": 4.092110670936178e-08, + "loss": 0.0111, + "step": 117320 + }, + { + "epoch": 34.13732906604597, + "grad_norm": 0.11824709922075272, + "learning_rate": 4.0616546121803874e-08, + "loss": 0.0077, + "step": 117330 + }, + { + "epoch": 34.14023858015712, + "grad_norm": 0.10081804543733597, + "learning_rate": 4.031312160069689e-08, + "loss": 0.0119, + "step": 117340 + }, + { + "epoch": 34.143148094268255, + "grad_norm": 0.09655119478702545, + "learning_rate": 4.0010833169082384e-08, + "loss": 0.0094, + "step": 117350 + }, + { + "epoch": 34.1460576083794, + "grad_norm": 0.16364642977714539, + "learning_rate": 3.9709680849916995e-08, + "loss": 0.011, + "step": 117360 + }, + { + "epoch": 34.148967122490546, + "grad_norm": 0.11534536629915237, + "learning_rate": 3.94096646660741e-08, + "loss": 0.0094, + "step": 117370 + }, + { + "epoch": 34.151876636601685, + "grad_norm": 0.08805130422115326, + "learning_rate": 3.9110784640333796e-08, + "loss": 0.0101, + "step": 117380 + }, + { + "epoch": 34.15478615071283, + "grad_norm": 0.11790336668491364, + "learning_rate": 3.881304079539627e-08, + "loss": 0.0107, + "step": 117390 + }, + { + "epoch": 34.157695664823976, + "grad_norm": 0.12371298670768738, + "learning_rate": 3.851643315387343e-08, + "loss": 0.009, + "step": 117400 + }, + { + "epoch": 34.16060517893512, + "grad_norm": 0.08400320261716843, + "learning_rate": 3.822096173829059e-08, + "loss": 0.0092, + "step": 117410 + }, + { + "epoch": 34.16351469304626, + "grad_norm": 0.09926670044660568, + "learning_rate": 3.7926626571086474e-08, + "loss": 0.0104, + "step": 117420 + }, + { + "epoch": 34.166424207157405, + "grad_norm": 0.10164133459329605, + "learning_rate": 3.763342767461486e-08, + "loss": 0.0104, + "step": 117430 + }, + { + "epoch": 34.16933372126855, + "grad_norm": 0.15864846110343933, + "learning_rate": 3.734136507113961e-08, + "loss": 0.0099, + "step": 117440 + }, + { + "epoch": 34.17224323537969, + "grad_norm": 0.11817216128110886, + "learning_rate": 3.705043878284464e-08, + "loss": 0.0108, + "step": 117450 + }, + { + "epoch": 34.175152749490834, + "grad_norm": 0.09451248496770859, + "learning_rate": 3.676064883182062e-08, + "loss": 0.0091, + "step": 117460 + }, + { + "epoch": 34.17806226360198, + "grad_norm": 0.13426122069358826, + "learning_rate": 3.6471995240076605e-08, + "loss": 0.009, + "step": 117470 + }, + { + "epoch": 34.180971777713125, + "grad_norm": 0.1258108764886856, + "learning_rate": 3.6184478029533394e-08, + "loss": 0.0096, + "step": 117480 + }, + { + "epoch": 34.183881291824264, + "grad_norm": 0.09399426728487015, + "learning_rate": 3.589809722202853e-08, + "loss": 0.0071, + "step": 117490 + }, + { + "epoch": 34.18679080593541, + "grad_norm": 0.09082839637994766, + "learning_rate": 3.561285283930626e-08, + "loss": 0.0105, + "step": 117500 + }, + { + "epoch": 34.189700320046555, + "grad_norm": 0.1567094326019287, + "learning_rate": 3.5328744903032614e-08, + "loss": 0.0104, + "step": 117510 + }, + { + "epoch": 34.19260983415769, + "grad_norm": 0.10562194138765335, + "learning_rate": 3.504577343478366e-08, + "loss": 0.0109, + "step": 117520 + }, + { + "epoch": 34.19551934826884, + "grad_norm": 0.11113633215427399, + "learning_rate": 3.47639384560472e-08, + "loss": 0.0124, + "step": 117530 + }, + { + "epoch": 34.198428862379984, + "grad_norm": 0.12058890610933304, + "learning_rate": 3.448323998822611e-08, + "loss": 0.0095, + "step": 117540 + }, + { + "epoch": 34.20133837649112, + "grad_norm": 0.06785599887371063, + "learning_rate": 3.420367805264002e-08, + "loss": 0.0103, + "step": 117550 + }, + { + "epoch": 34.20424789060227, + "grad_norm": 0.09618471562862396, + "learning_rate": 3.392525267052027e-08, + "loss": 0.0092, + "step": 117560 + }, + { + "epoch": 34.20715740471341, + "grad_norm": 0.1349690705537796, + "learning_rate": 3.3647963863009943e-08, + "loss": 0.0121, + "step": 117570 + }, + { + "epoch": 34.21006691882456, + "grad_norm": 0.12566038966178894, + "learning_rate": 3.3371811651165543e-08, + "loss": 0.0114, + "step": 117580 + }, + { + "epoch": 34.2129764329357, + "grad_norm": 0.10836420953273773, + "learning_rate": 3.309679605596194e-08, + "loss": 0.0106, + "step": 117590 + }, + { + "epoch": 34.21588594704684, + "grad_norm": 0.12892164289951324, + "learning_rate": 3.282291709828411e-08, + "loss": 0.0104, + "step": 117600 + }, + { + "epoch": 34.21879546115799, + "grad_norm": 0.13935622572898865, + "learning_rate": 3.255017479893041e-08, + "loss": 0.0093, + "step": 117610 + }, + { + "epoch": 34.22170497526913, + "grad_norm": 0.09380965679883957, + "learning_rate": 3.2278569178614273e-08, + "loss": 0.0099, + "step": 117620 + }, + { + "epoch": 34.22461448938027, + "grad_norm": 0.09493963420391083, + "learning_rate": 3.2008100257962524e-08, + "loss": 0.0123, + "step": 117630 + }, + { + "epoch": 34.22752400349142, + "grad_norm": 0.08989895135164261, + "learning_rate": 3.1738768057517073e-08, + "loss": 0.0096, + "step": 117640 + }, + { + "epoch": 34.23043351760256, + "grad_norm": 0.11424672603607178, + "learning_rate": 3.147057259772823e-08, + "loss": 0.011, + "step": 117650 + }, + { + "epoch": 34.2333430317137, + "grad_norm": 0.1675957590341568, + "learning_rate": 3.120351389896636e-08, + "loss": 0.0105, + "step": 117660 + }, + { + "epoch": 34.23625254582485, + "grad_norm": 0.13017885386943817, + "learning_rate": 3.0937591981511916e-08, + "loss": 0.0117, + "step": 117670 + }, + { + "epoch": 34.23916205993599, + "grad_norm": 0.09103302657604218, + "learning_rate": 3.06728068655604e-08, + "loss": 0.0117, + "step": 117680 + }, + { + "epoch": 34.24207157404713, + "grad_norm": 0.0829547718167305, + "learning_rate": 3.0409158571219063e-08, + "loss": 0.0092, + "step": 117690 + }, + { + "epoch": 34.244981088158276, + "grad_norm": 0.09096657484769821, + "learning_rate": 3.0146647118513556e-08, + "loss": 0.01, + "step": 117700 + }, + { + "epoch": 34.24789060226942, + "grad_norm": 0.08313652873039246, + "learning_rate": 2.988527252737627e-08, + "loss": 0.0099, + "step": 117710 + }, + { + "epoch": 34.25080011638057, + "grad_norm": 0.1180175319314003, + "learning_rate": 2.9625034817657993e-08, + "loss": 0.0109, + "step": 117720 + }, + { + "epoch": 34.253709630491706, + "grad_norm": 0.093678317964077, + "learning_rate": 2.936593400912457e-08, + "loss": 0.0095, + "step": 117730 + }, + { + "epoch": 34.25661914460285, + "grad_norm": 0.13519267737865448, + "learning_rate": 2.9107970121448612e-08, + "loss": 0.0107, + "step": 117740 + }, + { + "epoch": 34.259528658714, + "grad_norm": 0.09447869658470154, + "learning_rate": 2.8851143174224437e-08, + "loss": 0.0082, + "step": 117750 + }, + { + "epoch": 34.262438172825135, + "grad_norm": 0.07271480560302734, + "learning_rate": 2.8595453186954777e-08, + "loss": 0.0101, + "step": 117760 + }, + { + "epoch": 34.26534768693628, + "grad_norm": 0.1576179713010788, + "learning_rate": 2.8340900179057438e-08, + "loss": 0.0118, + "step": 117770 + }, + { + "epoch": 34.268257201047426, + "grad_norm": 0.10761675983667374, + "learning_rate": 2.808748416986362e-08, + "loss": 0.0104, + "step": 117780 + }, + { + "epoch": 34.27116671515857, + "grad_norm": 0.06409718841314316, + "learning_rate": 2.783520517862126e-08, + "loss": 0.0104, + "step": 117790 + }, + { + "epoch": 34.27407622926971, + "grad_norm": 0.08117791265249252, + "learning_rate": 2.7584063224485034e-08, + "loss": 0.0103, + "step": 117800 + }, + { + "epoch": 34.276985743380855, + "grad_norm": 0.11177060008049011, + "learning_rate": 2.7334058326531354e-08, + "loss": 0.0104, + "step": 117810 + }, + { + "epoch": 34.279895257492, + "grad_norm": 0.12478521466255188, + "learning_rate": 2.7085190503745027e-08, + "loss": 0.0089, + "step": 117820 + }, + { + "epoch": 34.28280477160314, + "grad_norm": 0.145394429564476, + "learning_rate": 2.683745977502594e-08, + "loss": 0.0087, + "step": 117830 + }, + { + "epoch": 34.285714285714285, + "grad_norm": 0.07797399163246155, + "learning_rate": 2.659086615918571e-08, + "loss": 0.0088, + "step": 117840 + }, + { + "epoch": 34.28862379982543, + "grad_norm": 0.09328898042440414, + "learning_rate": 2.6345409674956024e-08, + "loss": 0.0105, + "step": 117850 + }, + { + "epoch": 34.291533313936576, + "grad_norm": 0.11104235798120499, + "learning_rate": 2.610109034097363e-08, + "loss": 0.0108, + "step": 117860 + }, + { + "epoch": 34.294442828047714, + "grad_norm": 0.11560434103012085, + "learning_rate": 2.5857908175793697e-08, + "loss": 0.0115, + "step": 117870 + }, + { + "epoch": 34.29735234215886, + "grad_norm": 0.13893350958824158, + "learning_rate": 2.5615863197884782e-08, + "loss": 0.0096, + "step": 117880 + }, + { + "epoch": 34.300261856270005, + "grad_norm": 0.1430227905511856, + "learning_rate": 2.537495542563051e-08, + "loss": 0.0118, + "step": 117890 + }, + { + "epoch": 34.30317137038114, + "grad_norm": 0.13175611197948456, + "learning_rate": 2.5135184877322914e-08, + "loss": 0.0102, + "step": 117900 + }, + { + "epoch": 34.30608088449229, + "grad_norm": 0.12965530157089233, + "learning_rate": 2.4896551571172432e-08, + "loss": 0.0117, + "step": 117910 + }, + { + "epoch": 34.308990398603434, + "grad_norm": 0.06797997653484344, + "learning_rate": 2.4659055525302898e-08, + "loss": 0.0092, + "step": 117920 + }, + { + "epoch": 34.31189991271458, + "grad_norm": 0.09895024448633194, + "learning_rate": 2.442269675774822e-08, + "loss": 0.0101, + "step": 117930 + }, + { + "epoch": 34.31480942682572, + "grad_norm": 0.07761017978191376, + "learning_rate": 2.4187475286460702e-08, + "loss": 0.0101, + "step": 117940 + }, + { + "epoch": 34.317718940936864, + "grad_norm": 0.1276521235704422, + "learning_rate": 2.395339112930106e-08, + "loss": 0.0084, + "step": 117950 + }, + { + "epoch": 34.32062845504801, + "grad_norm": 0.08563243597745895, + "learning_rate": 2.372044430405007e-08, + "loss": 0.0102, + "step": 117960 + }, + { + "epoch": 34.32353796915915, + "grad_norm": 0.10490110516548157, + "learning_rate": 2.348863482839525e-08, + "loss": 0.0113, + "step": 117970 + }, + { + "epoch": 34.32644748327029, + "grad_norm": 0.1698862910270691, + "learning_rate": 2.3257962719942516e-08, + "loss": 0.0088, + "step": 117980 + }, + { + "epoch": 34.32935699738144, + "grad_norm": 0.15196150541305542, + "learning_rate": 2.302842799620952e-08, + "loss": 0.0134, + "step": 117990 + }, + { + "epoch": 34.332266511492584, + "grad_norm": 0.1656128317117691, + "learning_rate": 2.2800030674627327e-08, + "loss": 0.0093, + "step": 118000 + }, + { + "epoch": 34.33517602560372, + "grad_norm": 0.06592484563589096, + "learning_rate": 2.2572770772543718e-08, + "loss": 0.0102, + "step": 118010 + }, + { + "epoch": 34.33808553971487, + "grad_norm": 0.1315012276172638, + "learning_rate": 2.234664830721489e-08, + "loss": 0.011, + "step": 118020 + }, + { + "epoch": 34.34099505382601, + "grad_norm": 0.09701064974069595, + "learning_rate": 2.2121663295813776e-08, + "loss": 0.0106, + "step": 118030 + }, + { + "epoch": 34.34390456793715, + "grad_norm": 0.058127790689468384, + "learning_rate": 2.189781575542671e-08, + "loss": 0.0086, + "step": 118040 + }, + { + "epoch": 34.3468140820483, + "grad_norm": 0.08422420173883438, + "learning_rate": 2.167510570305342e-08, + "loss": 0.0119, + "step": 118050 + }, + { + "epoch": 34.34972359615944, + "grad_norm": 0.13378971815109253, + "learning_rate": 2.145353315560872e-08, + "loss": 0.0117, + "step": 118060 + }, + { + "epoch": 34.35263311027059, + "grad_norm": 0.1317129135131836, + "learning_rate": 2.123309812991747e-08, + "loss": 0.0104, + "step": 118070 + }, + { + "epoch": 34.35554262438173, + "grad_norm": 0.10610689967870712, + "learning_rate": 2.1013800642722957e-08, + "loss": 0.0099, + "step": 118080 + }, + { + "epoch": 34.35845213849287, + "grad_norm": 0.10956978052854538, + "learning_rate": 2.079564071067519e-08, + "loss": 0.0098, + "step": 118090 + }, + { + "epoch": 34.36136165260402, + "grad_norm": 0.12094477564096451, + "learning_rate": 2.0578618350345912e-08, + "loss": 0.0103, + "step": 118100 + }, + { + "epoch": 34.364271166715156, + "grad_norm": 0.07987771183252335, + "learning_rate": 2.036273357821361e-08, + "loss": 0.0083, + "step": 118110 + }, + { + "epoch": 34.3671806808263, + "grad_norm": 0.10668838769197464, + "learning_rate": 2.0147986410675168e-08, + "loss": 0.0111, + "step": 118120 + }, + { + "epoch": 34.37009019493745, + "grad_norm": 0.08355999737977982, + "learning_rate": 1.9934376864039204e-08, + "loss": 0.0107, + "step": 118130 + }, + { + "epoch": 34.37299970904859, + "grad_norm": 0.12821514904499054, + "learning_rate": 1.9721904954527747e-08, + "loss": 0.0095, + "step": 118140 + }, + { + "epoch": 34.37590922315973, + "grad_norm": 0.09075549244880676, + "learning_rate": 1.951057069827622e-08, + "loss": 0.0095, + "step": 118150 + }, + { + "epoch": 34.378818737270876, + "grad_norm": 0.08650843054056168, + "learning_rate": 1.9300374111333452e-08, + "loss": 0.0091, + "step": 118160 + }, + { + "epoch": 34.38172825138202, + "grad_norm": 0.15410354733467102, + "learning_rate": 1.909131520966334e-08, + "loss": 0.011, + "step": 118170 + }, + { + "epoch": 34.38463776549316, + "grad_norm": 0.11082956194877625, + "learning_rate": 1.8883394009143186e-08, + "loss": 0.0104, + "step": 118180 + }, + { + "epoch": 34.387547279604306, + "grad_norm": 0.16346053779125214, + "learning_rate": 1.867661052556202e-08, + "loss": 0.0112, + "step": 118190 + }, + { + "epoch": 34.39045679371545, + "grad_norm": 0.12810876965522766, + "learning_rate": 1.8470964774623954e-08, + "loss": 0.0136, + "step": 118200 + }, + { + "epoch": 34.39336630782659, + "grad_norm": 0.09576006978750229, + "learning_rate": 1.8266456771948158e-08, + "loss": 0.0115, + "step": 118210 + }, + { + "epoch": 34.396275821937735, + "grad_norm": 0.06292126327753067, + "learning_rate": 1.806308653306221e-08, + "loss": 0.0104, + "step": 118220 + }, + { + "epoch": 34.39918533604888, + "grad_norm": 0.17538714408874512, + "learning_rate": 1.786085407341376e-08, + "loss": 0.0097, + "step": 118230 + }, + { + "epoch": 34.402094850160026, + "grad_norm": 0.10045831650495529, + "learning_rate": 1.7659759408358845e-08, + "loss": 0.01, + "step": 118240 + }, + { + "epoch": 34.405004364271164, + "grad_norm": 0.11893518269062042, + "learning_rate": 1.745980255317192e-08, + "loss": 0.0116, + "step": 118250 + }, + { + "epoch": 34.40791387838231, + "grad_norm": 0.1293673813343048, + "learning_rate": 1.7260983523035846e-08, + "loss": 0.0116, + "step": 118260 + }, + { + "epoch": 34.410823392493455, + "grad_norm": 0.09910473227500916, + "learning_rate": 1.70633023330502e-08, + "loss": 0.0104, + "step": 118270 + }, + { + "epoch": 34.413732906604594, + "grad_norm": 0.11084427684545517, + "learning_rate": 1.686675899822965e-08, + "loss": 0.0112, + "step": 118280 + }, + { + "epoch": 34.41664242071574, + "grad_norm": 0.0747825875878334, + "learning_rate": 1.6671353533497247e-08, + "loss": 0.0092, + "step": 118290 + }, + { + "epoch": 34.419551934826885, + "grad_norm": 0.06408549845218658, + "learning_rate": 1.647708595369446e-08, + "loss": 0.0109, + "step": 118300 + }, + { + "epoch": 34.42246144893803, + "grad_norm": 0.07297014445066452, + "learning_rate": 1.6283956273574484e-08, + "loss": 0.0105, + "step": 118310 + }, + { + "epoch": 34.42537096304917, + "grad_norm": 0.1530429869890213, + "learning_rate": 1.6091964507803926e-08, + "loss": 0.0097, + "step": 118320 + }, + { + "epoch": 34.428280477160314, + "grad_norm": 0.10838034003973007, + "learning_rate": 1.5901110670962783e-08, + "loss": 0.011, + "step": 118330 + }, + { + "epoch": 34.43118999127146, + "grad_norm": 0.0926162376999855, + "learning_rate": 1.5711394777546127e-08, + "loss": 0.0077, + "step": 118340 + }, + { + "epoch": 34.4340995053826, + "grad_norm": 0.08453333377838135, + "learning_rate": 1.552281684196244e-08, + "loss": 0.0108, + "step": 118350 + }, + { + "epoch": 34.43700901949374, + "grad_norm": 0.10396793484687805, + "learning_rate": 1.5335376878531927e-08, + "loss": 0.0114, + "step": 118360 + }, + { + "epoch": 34.43991853360489, + "grad_norm": 0.07732348144054413, + "learning_rate": 1.5149074901488202e-08, + "loss": 0.0092, + "step": 118370 + }, + { + "epoch": 34.442828047716034, + "grad_norm": 0.12018019706010818, + "learning_rate": 1.4963910924981616e-08, + "loss": 0.0107, + "step": 118380 + }, + { + "epoch": 34.44573756182717, + "grad_norm": 0.10822589695453644, + "learning_rate": 1.4779884963072587e-08, + "loss": 0.0103, + "step": 118390 + }, + { + "epoch": 34.44864707593832, + "grad_norm": 0.10365793853998184, + "learning_rate": 1.4596997029736603e-08, + "loss": 0.0095, + "step": 118400 + }, + { + "epoch": 34.451556590049464, + "grad_norm": 0.07404147833585739, + "learning_rate": 1.4415247138865883e-08, + "loss": 0.0125, + "step": 118410 + }, + { + "epoch": 34.4544661041606, + "grad_norm": 0.14231297373771667, + "learning_rate": 1.423463530425939e-08, + "loss": 0.0086, + "step": 118420 + }, + { + "epoch": 34.45737561827175, + "grad_norm": 0.10408905148506165, + "learning_rate": 1.4055161539636152e-08, + "loss": 0.0094, + "step": 118430 + }, + { + "epoch": 34.46028513238289, + "grad_norm": 0.11197537183761597, + "learning_rate": 1.3876825858623598e-08, + "loss": 0.0106, + "step": 118440 + }, + { + "epoch": 34.46319464649404, + "grad_norm": 0.13348592817783356, + "learning_rate": 1.3699628274767561e-08, + "loss": 0.0131, + "step": 118450 + }, + { + "epoch": 34.46610416060518, + "grad_norm": 0.08384132385253906, + "learning_rate": 1.352356880152228e-08, + "loss": 0.0097, + "step": 118460 + }, + { + "epoch": 34.46901367471632, + "grad_norm": 0.10205044597387314, + "learning_rate": 1.3348647452260387e-08, + "loss": 0.0101, + "step": 118470 + }, + { + "epoch": 34.47192318882747, + "grad_norm": 0.12525640428066254, + "learning_rate": 1.317486424026626e-08, + "loss": 0.0089, + "step": 118480 + }, + { + "epoch": 34.474832702938606, + "grad_norm": 0.09847959876060486, + "learning_rate": 1.3002219178736007e-08, + "loss": 0.0123, + "step": 118490 + }, + { + "epoch": 34.47774221704975, + "grad_norm": 0.08540048450231552, + "learning_rate": 1.283071228078081e-08, + "loss": 0.0099, + "step": 118500 + }, + { + "epoch": 34.4806517311609, + "grad_norm": 0.10365810245275497, + "learning_rate": 1.2660343559426913e-08, + "loss": 0.0094, + "step": 118510 + }, + { + "epoch": 34.48356124527204, + "grad_norm": 0.12010686099529266, + "learning_rate": 1.24911130276123e-08, + "loss": 0.0099, + "step": 118520 + }, + { + "epoch": 34.48647075938318, + "grad_norm": 0.20123255252838135, + "learning_rate": 1.2323020698190023e-08, + "loss": 0.0112, + "step": 118530 + }, + { + "epoch": 34.489380273494326, + "grad_norm": 0.10049042105674744, + "learning_rate": 1.2156066583923208e-08, + "loss": 0.0088, + "step": 118540 + }, + { + "epoch": 34.49228978760547, + "grad_norm": 0.2056250423192978, + "learning_rate": 1.1990250697490047e-08, + "loss": 0.011, + "step": 118550 + }, + { + "epoch": 34.49519930171661, + "grad_norm": 0.12318383157253265, + "learning_rate": 1.1825573051487127e-08, + "loss": 0.0094, + "step": 118560 + }, + { + "epoch": 34.498108815827756, + "grad_norm": 0.12447409331798553, + "learning_rate": 1.1662033658419447e-08, + "loss": 0.0107, + "step": 118570 + }, + { + "epoch": 34.5010183299389, + "grad_norm": 0.09476412087678909, + "learning_rate": 1.149963253070374e-08, + "loss": 0.0101, + "step": 118580 + }, + { + "epoch": 34.50392784405005, + "grad_norm": 0.10836581140756607, + "learning_rate": 1.1338369680676809e-08, + "loss": 0.0113, + "step": 118590 + }, + { + "epoch": 34.506837358161185, + "grad_norm": 0.10033656656742096, + "learning_rate": 1.1178245120583851e-08, + "loss": 0.0115, + "step": 118600 + }, + { + "epoch": 34.50974687227233, + "grad_norm": 0.09494254738092422, + "learning_rate": 1.1019258862585146e-08, + "loss": 0.0099, + "step": 118610 + }, + { + "epoch": 34.512656386383476, + "grad_norm": 0.07287470996379852, + "learning_rate": 1.0861410918756032e-08, + "loss": 0.0109, + "step": 118620 + }, + { + "epoch": 34.515565900494614, + "grad_norm": 0.11417483538389206, + "learning_rate": 1.070470130108192e-08, + "loss": 0.0095, + "step": 118630 + }, + { + "epoch": 34.51847541460576, + "grad_norm": 0.07777088135480881, + "learning_rate": 1.0549130021464959e-08, + "loss": 0.0099, + "step": 118640 + }, + { + "epoch": 34.521384928716905, + "grad_norm": 0.07565118372440338, + "learning_rate": 1.0394697091720695e-08, + "loss": 0.0114, + "step": 118650 + }, + { + "epoch": 34.52429444282805, + "grad_norm": 0.07155228406190872, + "learning_rate": 1.0241402523576416e-08, + "loss": 0.011, + "step": 118660 + }, + { + "epoch": 34.52720395693919, + "grad_norm": 0.055621493607759476, + "learning_rate": 1.0089246328672807e-08, + "loss": 0.0084, + "step": 118670 + }, + { + "epoch": 34.530113471050335, + "grad_norm": 0.07400696724653244, + "learning_rate": 9.938228518567294e-09, + "loss": 0.0089, + "step": 118680 + }, + { + "epoch": 34.53302298516148, + "grad_norm": 0.09509781002998352, + "learning_rate": 9.788349104725703e-09, + "loss": 0.0099, + "step": 118690 + }, + { + "epoch": 34.53593249927262, + "grad_norm": 0.097788006067276, + "learning_rate": 9.639608098533925e-09, + "loss": 0.0082, + "step": 118700 + }, + { + "epoch": 34.538842013383764, + "grad_norm": 0.20659074187278748, + "learning_rate": 9.49200551128626e-09, + "loss": 0.0121, + "step": 118710 + }, + { + "epoch": 34.54175152749491, + "grad_norm": 0.3662491738796234, + "learning_rate": 9.345541354190413e-09, + "loss": 0.0101, + "step": 118720 + }, + { + "epoch": 34.544661041606055, + "grad_norm": 0.11423538625240326, + "learning_rate": 9.200215638372477e-09, + "loss": 0.0114, + "step": 118730 + }, + { + "epoch": 34.54757055571719, + "grad_norm": 0.11399926245212555, + "learning_rate": 9.056028374866965e-09, + "loss": 0.0094, + "step": 118740 + }, + { + "epoch": 34.55048006982834, + "grad_norm": 0.08401496708393097, + "learning_rate": 8.912979574625113e-09, + "loss": 0.0089, + "step": 118750 + }, + { + "epoch": 34.553389583939484, + "grad_norm": 0.13674820959568024, + "learning_rate": 8.771069248508234e-09, + "loss": 0.0101, + "step": 118760 + }, + { + "epoch": 34.55629909805062, + "grad_norm": 0.08315252512693405, + "learning_rate": 8.630297407297704e-09, + "loss": 0.0091, + "step": 118770 + }, + { + "epoch": 34.55920861216177, + "grad_norm": 0.0846276804804802, + "learning_rate": 8.490664061679977e-09, + "loss": 0.0102, + "step": 118780 + }, + { + "epoch": 34.562118126272914, + "grad_norm": 0.1452612727880478, + "learning_rate": 8.352169222259897e-09, + "loss": 0.0095, + "step": 118790 + }, + { + "epoch": 34.56502764038406, + "grad_norm": 0.11144470423460007, + "learning_rate": 8.214812899559054e-09, + "loss": 0.01, + "step": 118800 + }, + { + "epoch": 34.5679371544952, + "grad_norm": 0.07877112925052643, + "learning_rate": 8.078595104004106e-09, + "loss": 0.0105, + "step": 118810 + }, + { + "epoch": 34.57084666860634, + "grad_norm": 0.0921756848692894, + "learning_rate": 7.943515845941773e-09, + "loss": 0.0107, + "step": 118820 + }, + { + "epoch": 34.57375618271749, + "grad_norm": 0.11763215065002441, + "learning_rate": 7.80957513563052e-09, + "loss": 0.0107, + "step": 118830 + }, + { + "epoch": 34.57666569682863, + "grad_norm": 0.15027830004692078, + "learning_rate": 7.676772983243874e-09, + "loss": 0.0115, + "step": 118840 + }, + { + "epoch": 34.57957521093977, + "grad_norm": 0.06850502640008926, + "learning_rate": 7.545109398863769e-09, + "loss": 0.0089, + "step": 118850 + }, + { + "epoch": 34.58248472505092, + "grad_norm": 0.10216091573238373, + "learning_rate": 7.414584392490542e-09, + "loss": 0.0085, + "step": 118860 + }, + { + "epoch": 34.585394239162056, + "grad_norm": 0.10604722797870636, + "learning_rate": 7.285197974039592e-09, + "loss": 0.0101, + "step": 118870 + }, + { + "epoch": 34.5883037532732, + "grad_norm": 0.1198640838265419, + "learning_rate": 7.156950153333064e-09, + "loss": 0.0107, + "step": 118880 + }, + { + "epoch": 34.59121326738435, + "grad_norm": 0.11040479689836502, + "learning_rate": 7.029840940111498e-09, + "loss": 0.0091, + "step": 118890 + }, + { + "epoch": 34.59412278149549, + "grad_norm": 0.13447979092597961, + "learning_rate": 6.90387034402884e-09, + "loss": 0.0092, + "step": 118900 + }, + { + "epoch": 34.59703229560663, + "grad_norm": 0.10018313676118851, + "learning_rate": 6.779038374652435e-09, + "loss": 0.0114, + "step": 118910 + }, + { + "epoch": 34.59994180971778, + "grad_norm": 0.10307541489601135, + "learning_rate": 6.655345041459704e-09, + "loss": 0.0095, + "step": 118920 + }, + { + "epoch": 34.60285132382892, + "grad_norm": 0.0813726857304573, + "learning_rate": 6.5327903538464625e-09, + "loss": 0.0103, + "step": 118930 + }, + { + "epoch": 34.60576083794006, + "grad_norm": 0.08839832246303558, + "learning_rate": 6.411374321120267e-09, + "loss": 0.0112, + "step": 118940 + }, + { + "epoch": 34.608670352051206, + "grad_norm": 0.1343439221382141, + "learning_rate": 6.2910969525020735e-09, + "loss": 0.0084, + "step": 118950 + }, + { + "epoch": 34.61157986616235, + "grad_norm": 0.1306726634502411, + "learning_rate": 6.1719582571229115e-09, + "loss": 0.0095, + "step": 118960 + }, + { + "epoch": 34.6144893802735, + "grad_norm": 0.08313843607902527, + "learning_rate": 6.0539582440338745e-09, + "loss": 0.0098, + "step": 118970 + }, + { + "epoch": 34.617398894384635, + "grad_norm": 0.10071597248315811, + "learning_rate": 5.937096922196128e-09, + "loss": 0.0117, + "step": 118980 + }, + { + "epoch": 34.62030840849578, + "grad_norm": 0.09036128968000412, + "learning_rate": 5.821374300482574e-09, + "loss": 0.0094, + "step": 118990 + }, + { + "epoch": 34.623217922606926, + "grad_norm": 0.11580657958984375, + "learning_rate": 5.706790387682848e-09, + "loss": 0.009, + "step": 119000 + }, + { + "epoch": 34.626127436718065, + "grad_norm": 0.11457178741693497, + "learning_rate": 5.593345192498322e-09, + "loss": 0.0108, + "step": 119010 + }, + { + "epoch": 34.62903695082921, + "grad_norm": 0.083550825715065, + "learning_rate": 5.481038723545439e-09, + "loss": 0.0091, + "step": 119020 + }, + { + "epoch": 34.631946464940356, + "grad_norm": 0.11705878376960754, + "learning_rate": 5.369870989352377e-09, + "loss": 0.0106, + "step": 119030 + }, + { + "epoch": 34.6348559790515, + "grad_norm": 0.0785648450255394, + "learning_rate": 5.259841998362381e-09, + "loss": 0.0091, + "step": 119040 + }, + { + "epoch": 34.63776549316264, + "grad_norm": 0.06956597417593002, + "learning_rate": 5.150951758928768e-09, + "loss": 0.0089, + "step": 119050 + }, + { + "epoch": 34.640675007273785, + "grad_norm": 0.12715494632720947, + "learning_rate": 5.043200279324922e-09, + "loss": 0.0102, + "step": 119060 + }, + { + "epoch": 34.64358452138493, + "grad_norm": 0.10495741665363312, + "learning_rate": 4.936587567730965e-09, + "loss": 0.0104, + "step": 119070 + }, + { + "epoch": 34.64649403549607, + "grad_norm": 0.10611774772405624, + "learning_rate": 4.831113632243755e-09, + "loss": 0.0098, + "step": 119080 + }, + { + "epoch": 34.649403549607214, + "grad_norm": 0.14300262928009033, + "learning_rate": 4.726778480875216e-09, + "loss": 0.0098, + "step": 119090 + }, + { + "epoch": 34.65231306371836, + "grad_norm": 0.14386862516403198, + "learning_rate": 4.6235821215473425e-09, + "loss": 0.0109, + "step": 119100 + }, + { + "epoch": 34.655222577829505, + "grad_norm": 0.1282382756471634, + "learning_rate": 4.521524562097201e-09, + "loss": 0.0095, + "step": 119110 + }, + { + "epoch": 34.658132091940644, + "grad_norm": 0.09122148901224136, + "learning_rate": 4.420605810275258e-09, + "loss": 0.0097, + "step": 119120 + }, + { + "epoch": 34.66104160605179, + "grad_norm": 0.08903498947620392, + "learning_rate": 4.320825873747048e-09, + "loss": 0.0115, + "step": 119130 + }, + { + "epoch": 34.663951120162935, + "grad_norm": 0.11634386330842972, + "learning_rate": 4.222184760088177e-09, + "loss": 0.0126, + "step": 119140 + }, + { + "epoch": 34.66686063427407, + "grad_norm": 0.09226131439208984, + "learning_rate": 4.124682476792652e-09, + "loss": 0.0105, + "step": 119150 + }, + { + "epoch": 34.66977014838522, + "grad_norm": 0.08647021651268005, + "learning_rate": 4.028319031261218e-09, + "loss": 0.0096, + "step": 119160 + }, + { + "epoch": 34.672679662496364, + "grad_norm": 0.08562306314706802, + "learning_rate": 3.933094430816353e-09, + "loss": 0.0093, + "step": 119170 + }, + { + "epoch": 34.67558917660751, + "grad_norm": 0.08379307389259338, + "learning_rate": 3.8390086826856076e-09, + "loss": 0.0101, + "step": 119180 + }, + { + "epoch": 34.67849869071865, + "grad_norm": 0.12314529716968536, + "learning_rate": 3.746061794016597e-09, + "loss": 0.0101, + "step": 119190 + }, + { + "epoch": 34.68140820482979, + "grad_norm": 0.09673997759819031, + "learning_rate": 3.654253771868676e-09, + "loss": 0.0108, + "step": 119200 + }, + { + "epoch": 34.68431771894094, + "grad_norm": 0.08222205191850662, + "learning_rate": 3.563584623211269e-09, + "loss": 0.0095, + "step": 119210 + }, + { + "epoch": 34.68722723305208, + "grad_norm": 0.10926046967506409, + "learning_rate": 3.474054354932199e-09, + "loss": 0.009, + "step": 119220 + }, + { + "epoch": 34.69013674716322, + "grad_norm": 0.07038076967000961, + "learning_rate": 3.3856629738310273e-09, + "loss": 0.0099, + "step": 119230 + }, + { + "epoch": 34.69304626127437, + "grad_norm": 0.09532737731933594, + "learning_rate": 3.298410486619052e-09, + "loss": 0.0106, + "step": 119240 + }, + { + "epoch": 34.695955775385514, + "grad_norm": 0.09032963216304779, + "learning_rate": 3.212296899924305e-09, + "loss": 0.0098, + "step": 119250 + }, + { + "epoch": 34.69886528949665, + "grad_norm": 0.17020151019096375, + "learning_rate": 3.127322220283224e-09, + "loss": 0.0114, + "step": 119260 + }, + { + "epoch": 34.7017748036078, + "grad_norm": 0.09876697510480881, + "learning_rate": 3.043486454152311e-09, + "loss": 0.0095, + "step": 119270 + }, + { + "epoch": 34.70468431771894, + "grad_norm": 0.10236819833517075, + "learning_rate": 2.96078960789814e-09, + "loss": 0.0086, + "step": 119280 + }, + { + "epoch": 34.70759383183008, + "grad_norm": 0.08062794059515, + "learning_rate": 2.879231687799022e-09, + "loss": 0.0114, + "step": 119290 + }, + { + "epoch": 34.71050334594123, + "grad_norm": 0.08825766295194626, + "learning_rate": 2.798812700050002e-09, + "loss": 0.0113, + "step": 119300 + }, + { + "epoch": 34.71341286005237, + "grad_norm": 0.0811181589961052, + "learning_rate": 2.7195326507595264e-09, + "loss": 0.0101, + "step": 119310 + }, + { + "epoch": 34.71632237416352, + "grad_norm": 0.10285723954439163, + "learning_rate": 2.6413915459461146e-09, + "loss": 0.0092, + "step": 119320 + }, + { + "epoch": 34.719231888274656, + "grad_norm": 0.0981115847826004, + "learning_rate": 2.5643893915450187e-09, + "loss": 0.0102, + "step": 119330 + }, + { + "epoch": 34.7221414023858, + "grad_norm": 0.11174961924552917, + "learning_rate": 2.4885261934032287e-09, + "loss": 0.0121, + "step": 119340 + }, + { + "epoch": 34.72505091649695, + "grad_norm": 0.10954218357801437, + "learning_rate": 2.413801957284467e-09, + "loss": 0.0085, + "step": 119350 + }, + { + "epoch": 34.727960430608086, + "grad_norm": 0.11883050203323364, + "learning_rate": 2.3402166888608635e-09, + "loss": 0.0093, + "step": 119360 + }, + { + "epoch": 34.73086994471923, + "grad_norm": 0.08560995012521744, + "learning_rate": 2.2677703937229453e-09, + "loss": 0.0107, + "step": 119370 + }, + { + "epoch": 34.73377945883038, + "grad_norm": 0.06578007340431213, + "learning_rate": 2.1964630773713136e-09, + "loss": 0.0127, + "step": 119380 + }, + { + "epoch": 34.73668897294152, + "grad_norm": 0.13476215302944183, + "learning_rate": 2.1262947452199697e-09, + "loss": 0.0101, + "step": 119390 + }, + { + "epoch": 34.73959848705266, + "grad_norm": 0.08878269791603088, + "learning_rate": 2.057265402601316e-09, + "loss": 0.011, + "step": 119400 + }, + { + "epoch": 34.742508001163806, + "grad_norm": 0.08242614567279816, + "learning_rate": 1.989375054754494e-09, + "loss": 0.0104, + "step": 119410 + }, + { + "epoch": 34.74541751527495, + "grad_norm": 0.1021231859922409, + "learning_rate": 1.92262370683538e-09, + "loss": 0.0106, + "step": 119420 + }, + { + "epoch": 34.74832702938609, + "grad_norm": 0.23770035803318024, + "learning_rate": 1.8570113639149177e-09, + "loss": 0.01, + "step": 119430 + }, + { + "epoch": 34.751236543497235, + "grad_norm": 0.1370684653520584, + "learning_rate": 1.7925380309757878e-09, + "loss": 0.0097, + "step": 119440 + }, + { + "epoch": 34.75414605760838, + "grad_norm": 0.1272241622209549, + "learning_rate": 1.7292037129140737e-09, + "loss": 0.0129, + "step": 119450 + }, + { + "epoch": 34.757055571719526, + "grad_norm": 0.08738737553358078, + "learning_rate": 1.6670084145392617e-09, + "loss": 0.0099, + "step": 119460 + }, + { + "epoch": 34.759965085830665, + "grad_norm": 0.08117683976888657, + "learning_rate": 1.6059521405742406e-09, + "loss": 0.0107, + "step": 119470 + }, + { + "epoch": 34.76287459994181, + "grad_norm": 0.14487554132938385, + "learning_rate": 1.546034895656967e-09, + "loss": 0.0104, + "step": 119480 + }, + { + "epoch": 34.765784114052956, + "grad_norm": 0.08285524696111679, + "learning_rate": 1.487256684337135e-09, + "loss": 0.009, + "step": 119490 + }, + { + "epoch": 34.768693628164094, + "grad_norm": 0.07557661086320877, + "learning_rate": 1.4296175110778409e-09, + "loss": 0.0094, + "step": 119500 + }, + { + "epoch": 34.77160314227524, + "grad_norm": 0.088335782289505, + "learning_rate": 1.3731173802572495e-09, + "loss": 0.0109, + "step": 119510 + }, + { + "epoch": 34.774512656386385, + "grad_norm": 0.13575346767902374, + "learning_rate": 1.3177562961669276e-09, + "loss": 0.0103, + "step": 119520 + }, + { + "epoch": 34.77742217049753, + "grad_norm": 0.13809922337532043, + "learning_rate": 1.2635342630085144e-09, + "loss": 0.0094, + "step": 119530 + }, + { + "epoch": 34.78033168460867, + "grad_norm": 0.06719715148210526, + "learning_rate": 1.210451284903713e-09, + "loss": 0.011, + "step": 119540 + }, + { + "epoch": 34.783241198719814, + "grad_norm": 0.13888758420944214, + "learning_rate": 1.1585073658809675e-09, + "loss": 0.0112, + "step": 119550 + }, + { + "epoch": 34.78615071283096, + "grad_norm": 0.1405944973230362, + "learning_rate": 1.1077025098854554e-09, + "loss": 0.0104, + "step": 119560 + }, + { + "epoch": 34.7890602269421, + "grad_norm": 0.09846333414316177, + "learning_rate": 1.058036720775757e-09, + "loss": 0.0122, + "step": 119570 + }, + { + "epoch": 34.791969741053244, + "grad_norm": 0.10489160567522049, + "learning_rate": 1.00951000232552e-09, + "loss": 0.0086, + "step": 119580 + }, + { + "epoch": 34.79487925516439, + "grad_norm": 0.08173870295286179, + "learning_rate": 9.621223582167993e-10, + "loss": 0.0103, + "step": 119590 + }, + { + "epoch": 34.79778876927553, + "grad_norm": 0.0826789066195488, + "learning_rate": 9.158737920500481e-10, + "loss": 0.0092, + "step": 119600 + }, + { + "epoch": 34.80069828338667, + "grad_norm": 0.10311046987771988, + "learning_rate": 8.707643073357918e-10, + "loss": 0.0112, + "step": 119610 + }, + { + "epoch": 34.80360779749782, + "grad_norm": 0.08247760683298111, + "learning_rate": 8.26793907502954e-10, + "loss": 0.0118, + "step": 119620 + }, + { + "epoch": 34.806517311608964, + "grad_norm": 0.08376117050647736, + "learning_rate": 7.839625958905306e-10, + "loss": 0.0101, + "step": 119630 + }, + { + "epoch": 34.8094268257201, + "grad_norm": 0.10634606331586838, + "learning_rate": 7.422703757475891e-10, + "loss": 0.0098, + "step": 119640 + }, + { + "epoch": 34.81233633983125, + "grad_norm": 0.05403178930282593, + "learning_rate": 7.017172502432612e-10, + "loss": 0.0098, + "step": 119650 + }, + { + "epoch": 34.81524585394239, + "grad_norm": 0.05598849803209305, + "learning_rate": 6.623032224584158e-10, + "loss": 0.0097, + "step": 119660 + }, + { + "epoch": 34.81815536805353, + "grad_norm": 0.12693478167057037, + "learning_rate": 6.240282953823284e-10, + "loss": 0.0121, + "step": 119670 + }, + { + "epoch": 34.82106488216468, + "grad_norm": 0.13713228702545166, + "learning_rate": 5.868924719243385e-10, + "loss": 0.0108, + "step": 119680 + }, + { + "epoch": 34.82397439627582, + "grad_norm": 0.06625296175479889, + "learning_rate": 5.508957549055227e-10, + "loss": 0.0087, + "step": 119690 + }, + { + "epoch": 34.82688391038697, + "grad_norm": 0.0971950963139534, + "learning_rate": 5.16038147058695e-10, + "loss": 0.0093, + "step": 119700 + }, + { + "epoch": 34.829793424498106, + "grad_norm": 0.07810936123132706, + "learning_rate": 4.82319651030072e-10, + "loss": 0.0097, + "step": 119710 + }, + { + "epoch": 34.83270293860925, + "grad_norm": 0.12055648118257523, + "learning_rate": 4.4974026938093824e-10, + "loss": 0.0115, + "step": 119720 + }, + { + "epoch": 34.8356124527204, + "grad_norm": 0.11349856108427048, + "learning_rate": 4.183000045859808e-10, + "loss": 0.0097, + "step": 119730 + }, + { + "epoch": 34.838521966831536, + "grad_norm": 0.16184769570827484, + "learning_rate": 3.8799885903162415e-10, + "loss": 0.0128, + "step": 119740 + }, + { + "epoch": 34.84143148094268, + "grad_norm": 0.09972194582223892, + "learning_rate": 3.5883683502102583e-10, + "loss": 0.011, + "step": 119750 + }, + { + "epoch": 34.84434099505383, + "grad_norm": 0.20967313647270203, + "learning_rate": 3.308139347674155e-10, + "loss": 0.0129, + "step": 119760 + }, + { + "epoch": 34.84725050916497, + "grad_norm": 0.0812348946928978, + "learning_rate": 3.039301603990907e-10, + "loss": 0.0104, + "step": 119770 + }, + { + "epoch": 34.85016002327611, + "grad_norm": 0.10218928009271622, + "learning_rate": 2.7818551395775163e-10, + "loss": 0.0079, + "step": 119780 + }, + { + "epoch": 34.853069537387256, + "grad_norm": 0.07040470093488693, + "learning_rate": 2.535799973985009e-10, + "loss": 0.0097, + "step": 119790 + }, + { + "epoch": 34.8559790514984, + "grad_norm": 0.1281152367591858, + "learning_rate": 2.3011361258984397e-10, + "loss": 0.0086, + "step": 119800 + }, + { + "epoch": 34.85888856560954, + "grad_norm": 0.08776123821735382, + "learning_rate": 2.077863613153541e-10, + "loss": 0.011, + "step": 119810 + }, + { + "epoch": 34.861798079720685, + "grad_norm": 0.11854979395866394, + "learning_rate": 1.8659824526867653e-10, + "loss": 0.0098, + "step": 119820 + }, + { + "epoch": 34.86470759383183, + "grad_norm": 0.06477639079093933, + "learning_rate": 1.6654926606018973e-10, + "loss": 0.0076, + "step": 119830 + }, + { + "epoch": 34.867617107942976, + "grad_norm": 0.1340726613998413, + "learning_rate": 1.4763942521200946e-10, + "loss": 0.0115, + "step": 119840 + }, + { + "epoch": 34.870526622054115, + "grad_norm": 0.11204170435667038, + "learning_rate": 1.2986872415965413e-10, + "loss": 0.0112, + "step": 119850 + }, + { + "epoch": 34.87343613616526, + "grad_norm": 0.10497820377349854, + "learning_rate": 1.1323716425371e-10, + "loss": 0.01, + "step": 119860 + }, + { + "epoch": 34.876345650276406, + "grad_norm": 0.08375784754753113, + "learning_rate": 9.77447467565007e-11, + "loss": 0.0088, + "step": 119870 + }, + { + "epoch": 34.879255164387544, + "grad_norm": 0.1111755520105362, + "learning_rate": 8.33914728454177e-11, + "loss": 0.0104, + "step": 119880 + }, + { + "epoch": 34.88216467849869, + "grad_norm": 0.082642562687397, + "learning_rate": 7.017734360958983e-11, + "loss": 0.0119, + "step": 119890 + }, + { + "epoch": 34.885074192609835, + "grad_norm": 0.0982419028878212, + "learning_rate": 5.810236005154845e-11, + "loss": 0.0089, + "step": 119900 + }, + { + "epoch": 34.88798370672098, + "grad_norm": 0.10877154767513275, + "learning_rate": 4.716652309222358e-11, + "loss": 0.0114, + "step": 119910 + }, + { + "epoch": 34.89089322083212, + "grad_norm": 0.06789495795965195, + "learning_rate": 3.7369833557621137e-11, + "loss": 0.0105, + "step": 119920 + }, + { + "epoch": 34.893802734943264, + "grad_norm": 0.08195552974939346, + "learning_rate": 2.8712292193810995e-11, + "loss": 0.012, + "step": 119930 + }, + { + "epoch": 34.89671224905441, + "grad_norm": 0.09670759737491608, + "learning_rate": 2.11938996586003e-11, + "loss": 0.0087, + "step": 119940 + }, + { + "epoch": 34.89962176316555, + "grad_norm": 0.10317841917276382, + "learning_rate": 1.4814656521533464e-11, + "loss": 0.01, + "step": 119950 + }, + { + "epoch": 34.902531277276694, + "grad_norm": 0.15020567178726196, + "learning_rate": 9.57456326888817e-12, + "loss": 0.0107, + "step": 119960 + }, + { + "epoch": 34.90544079138784, + "grad_norm": 0.1235247328877449, + "learning_rate": 5.4736202986793714e-12, + "loss": 0.009, + "step": 119970 + }, + { + "epoch": 34.908350305498985, + "grad_norm": 0.11598283797502518, + "learning_rate": 2.5118279189939585e-12, + "loss": 0.0098, + "step": 119980 + }, + { + "epoch": 34.91125981961012, + "grad_norm": 0.11540386825799942, + "learning_rate": 6.891863596480974e-13, + "loss": 0.0083, + "step": 119990 + }, + { + "epoch": 34.91416933372127, + "grad_norm": 0.09390421211719513, + "learning_rate": 5.69575553388546e-15, + "loss": 0.0119, + "step": 120000 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 20000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 64, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-20000/config.json b/checkpoint-20000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d7b77e4255263f3803a4b6f94ca6831e459b70 --- /dev/null +++ b/checkpoint-20000/config.json @@ -0,0 +1,64 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/checkpoint-20000/experiment_cfg/metadata.json b/checkpoint-20000/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..300eb070963337016e16465f50ba8351ea2eb25c --- /dev/null +++ b/checkpoint-20000/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.3139495849609375, + 0.4876587688922882, + 0.7196521759033203 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6996827125549316 + ], + "mean": [ + 2.749288365724937, + -2.0455216239909983, + 0.700532551020533 + ], + "std": [ + 1.6786295897046262, + 1.312897969434244, + 0.00133751758906258 + ], + "q01": [ + -1.4241907881148037, + -5.1716547935950885, + 0.6999670898768614 + ], + "q99": [ + 6.198111545831555, + -0.5873531334104489, + 0.7038833014242587 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.23084999587450247, + 0.6238431088581847 + ], + "std": [ + 0.0, + 0.0, + 0.6556404371644047, + 0.3572252105732042 + ], + "q01": [ + 0.0, + 0.0, + -0.9999999999999999, + 1.7482354593273349e-06 + ], + "q99": [ + 0.0, + 0.0, + 0.9999999999999999, + 0.9999999999999999 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8003877401351929, + 0.9829942584037781 + ], + "min": [ + -0.37471598386764526, + -0.8472502827644348, + -0.25070279836654663 + ], + "mean": [ + 0.28667332601863704, + -0.038315473170876704, + 0.45974658906777444 + ], + "std": [ + 0.17067058828305154, + 0.23569952037784195, + 0.21950702354181487 + ], + "q01": [ + 0.015172948963784228, + -0.44450502063220615, + 0.23314444103329338 + ], + "q99": [ + 0.5609069689572412, + 0.38454179128009547, + 0.7028102736473852 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9984455108642578, + 0.9434727430343628, + 0.9062229990959167 + ], + "min": [ + -0.9999930262565613, + -0.9988337755203247, + -0.9618149995803833, + 2.1872274658107926e-07 + ], + "mean": [ + -0.25672297401336, + 0.024425335806687216, + -0.08740977164483847, + 0.16608739644018397 + ], + "std": [ + 0.7932585208392383, + 0.3102260195580416, + 0.3772472576315148, + 0.17451959300158773 + ], + "q01": [ + -0.99379006987882, + -0.5478102050038828, + -0.6101777952971636, + 0.002274085014134748 + ], + "q99": [ + 0.8804856038896288, + 0.5910611092873037, + 0.5216058042993562, + 0.5231965974012255 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011436971835792065, + -0.05664053186774254 + ], + "mean": [ + 0.031651551516052555, + -0.03161481387920421 + ], + "std": [ + 0.013119769222217526, + 0.01306078225192402 + ], + "q01": [ + 0.0060999246446777336, + -0.04062601653964198 + ], + "q99": [ + 0.04054891621517283, + -0.0059163567113456345 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.008144896753718373, + -0.00018893135146078263, + -0.0008739062727221845, + 0.0 + ], + "std": [ + 0.11030045424364411, + 0.10148082570313594, + 0.08975698373467506, + 0.0 + ], + "q01": [ + -0.07287373067581518, + -0.09446994199569948, + -0.07702818259249572, + 0.0 + ], + "q99": [ + 0.04485485414407898, + 0.09626941605965177, + 0.07610665906122742, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9221974313425939 + ], + "std": [ + 0.38670194342612674 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.384693946883099 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.01976129923017491, + -0.020870631344490034, + -0.05993476517349181 + ], + "std": [ + 0.4347024990804053, + 0.4221662976569997, + 0.3845506843044066 + ], + "q01": [ + -0.8835792939767807, + -0.9280919251713778, + -0.783361071470956 + ], + "q99": [ + 0.7622084438449307, + 0.8625125252609077, + 0.7470974753250706 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.008089673068544335, + -0.026498254627927348, + 0.0029083551743059005 + ], + "std": [ + 0.11216734503733773, + 0.13206002613798629, + 0.12615870412692864 + ], + "q01": [ + -0.2814627972846672, + -0.4342386516115439, + -0.34489755628340574 + ], + "q99": [ + 0.31879353266939386, + 0.28411797683220563, + 0.3597718671669894 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.3824228874769045 + ], + "std": [ + 0.9240015096469786 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.5597272305813592 + ] + } + } + }, + "modalities": { + "video": { + "robot0_eye_in_hand": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_left": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_right": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": true, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + }, + "end_effector_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": true, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/checkpoint-20000/model-00001-of-00002.safetensors b/checkpoint-20000/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5b5bc2d5468872f30bdd6a47850f7395afd8f965 --- /dev/null +++ b/checkpoint-20000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:964ec9f4969b7e71db017b9126c64b9bf4a9306f1870efdbbf25cf7f51f9c898 +size 4999367032 diff --git a/checkpoint-20000/model-00002-of-00002.safetensors b/checkpoint-20000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7c56d5b4487fc043ed9fe64efc27b297de1a6403 --- /dev/null +++ b/checkpoint-20000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e53b363c2a99db26e9504c237c37bafb3a1a01a9344f32e00a9a9222b171f95 +size 2586705312 diff --git a/checkpoint-20000/model.safetensors.index.json b/checkpoint-20000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/checkpoint-20000/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-20000/optimizer.pt b/checkpoint-20000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..27e5acd95928ee1c0bb8eca35cb0672d03f20aa6 --- /dev/null +++ b/checkpoint-20000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c1ad69b4c4256a1038449335c525661deb594f75b745cf5f39ecbd6b7d1bd65 +size 8550719959 diff --git a/checkpoint-20000/rng_state_0.pth b/checkpoint-20000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..37472edc360704b36d38353d98755b0d9c1b6f0b --- /dev/null +++ b/checkpoint-20000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5aa4d1e41b7def26adc435d9a8b4b9a70eda76d5acfcc486f0180e1b52251911 +size 14917 diff --git a/checkpoint-20000/rng_state_1.pth b/checkpoint-20000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..ae42de01c56c82c02fc1b20d26db1b485f6a983c --- /dev/null +++ b/checkpoint-20000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cdd05d28bd7bdbeca2bbac4d3c176387394b207613dd2a6a5875f3576ee054ef +size 14917 diff --git a/checkpoint-20000/scheduler.pt b/checkpoint-20000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..4ab023769c9a080e741413a2434c4b63fde0c9c3 --- /dev/null +++ b/checkpoint-20000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:112a38aafc3f98b3fd755c0c92a466440b5e73462b3129b51add61f925ce2ad3 +size 1465 diff --git a/checkpoint-20000/trainer_state.json b/checkpoint-20000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..e7871b144e3c0ca23829e2e259bc620dabc30eb3 --- /dev/null +++ b/checkpoint-20000/trainer_state.json @@ -0,0 +1,14034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 5.819028222286878, + "eval_steps": 500, + "global_step": 20000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002909514111143439, + "grad_norm": 5.5908684730529785, + "learning_rate": 4.5e-08, + "loss": 0.5668, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.084368705749512, + "learning_rate": 9.5e-08, + "loss": 0.5781, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 5.755157947540283, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5759, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.174816608428955, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5202, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 6.199656963348389, + "learning_rate": 2.45e-07, + "loss": 0.5455, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.251872539520264, + "learning_rate": 2.95e-07, + "loss": 0.5507, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 5.606583595275879, + "learning_rate": 3.45e-07, + "loss": 0.5264, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.967868328094482, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.5166, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.92909574508667, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4578, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.306433200836182, + "learning_rate": 4.95e-07, + "loss": 0.4091, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.75846529006958, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4072, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.2030856609344482, + "learning_rate": 5.95e-07, + "loss": 0.2992, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.5756216049194336, + "learning_rate": 6.45e-07, + "loss": 0.2703, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.3420923948287964, + "learning_rate": 6.95e-07, + "loss": 0.2492, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 0.9191234111785889, + "learning_rate": 7.45e-07, + "loss": 0.2302, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.662022054195404, + "learning_rate": 7.95e-07, + "loss": 0.1993, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.704409658908844, + "learning_rate": 8.45e-07, + "loss": 0.1814, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.6104884147644043, + "learning_rate": 8.95e-07, + "loss": 0.1694, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.4955693781375885, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1674, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.707667887210846, + "learning_rate": 9.95e-07, + "loss": 0.1507, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.6977017521858215, + "learning_rate": 1.045e-06, + "loss": 0.1547, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.49085867404937744, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1536, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.6363996267318726, + "learning_rate": 1.145e-06, + "loss": 0.1565, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.4633055627346039, + "learning_rate": 1.195e-06, + "loss": 0.1641, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5155240297317505, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1513, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5815547108650208, + "learning_rate": 1.295e-06, + "loss": 0.1431, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.5865835547447205, + "learning_rate": 1.345e-06, + "loss": 0.1419, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.5872474312782288, + "learning_rate": 1.395e-06, + "loss": 0.1326, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.4897519052028656, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1326, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.46316635608673096, + "learning_rate": 1.495e-06, + "loss": 0.1442, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.5359047651290894, + "learning_rate": 1.545e-06, + "loss": 0.1394, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.5805708169937134, + "learning_rate": 1.595e-06, + "loss": 0.1231, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.5488415360450745, + "learning_rate": 1.645e-06, + "loss": 0.127, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.41755908727645874, + "learning_rate": 1.695e-06, + "loss": 0.1375, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.6258692145347595, + "learning_rate": 1.745e-06, + "loss": 0.1202, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.5017617344856262, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1317, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.710078239440918, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.46977895498275757, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1215, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.4802077114582062, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.116, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.6448793411254883, + "learning_rate": 1.995e-06, + "loss": 0.1172, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4394509494304657, + "learning_rate": 2.045e-06, + "loss": 0.1145, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.6547806859016418, + "learning_rate": 2.095e-06, + "loss": 0.1174, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.6954914331436157, + "learning_rate": 2.145e-06, + "loss": 0.1245, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.670100212097168, + "learning_rate": 2.195e-06, + "loss": 0.1133, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.5230806469917297, + "learning_rate": 2.245e-06, + "loss": 0.11, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.46809589862823486, + "learning_rate": 2.295e-06, + "loss": 0.1069, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.6178835034370422, + "learning_rate": 2.345e-06, + "loss": 0.1099, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.41567298769950867, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1072, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4968737065792084, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1093, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.46893924474716187, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1046, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.5041665434837341, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1014, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5833673477172852, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1095, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.5565881133079529, + "learning_rate": 2.645e-06, + "loss": 0.1057, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6213339567184448, + "learning_rate": 2.695e-06, + "loss": 0.1087, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.5638794898986816, + "learning_rate": 2.745e-06, + "loss": 0.099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.5594047904014587, + "learning_rate": 2.795e-06, + "loss": 0.1065, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5521039366722107, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1072, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.5013270378112793, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1045, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.44355231523513794, + "learning_rate": 2.945e-06, + "loss": 0.0985, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.5467342734336853, + "learning_rate": 2.995e-06, + "loss": 0.101, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5065611600875854, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0939, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.5587425827980042, + "learning_rate": 3.095e-06, + "loss": 0.0969, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5561433434486389, + "learning_rate": 3.145e-06, + "loss": 0.1009, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.5576680302619934, + "learning_rate": 3.195e-06, + "loss": 0.1006, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.7426344752311707, + "learning_rate": 3.245e-06, + "loss": 0.0959, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.64913409948349, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.0962, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.5271337628364563, + "learning_rate": 3.345e-06, + "loss": 0.0927, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.6992810368537903, + "learning_rate": 3.395e-06, + "loss": 0.0924, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.559481680393219, + "learning_rate": 3.445e-06, + "loss": 0.0886, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7303935885429382, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0918, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.45654281973838806, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0934, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.41098082065582275, + "learning_rate": 3.595e-06, + "loss": 0.0911, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.43506675958633423, + "learning_rate": 3.645e-06, + "loss": 0.0887, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5131197571754456, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0983, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.4571124017238617, + "learning_rate": 3.745e-06, + "loss": 0.0928, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.6749465465545654, + "learning_rate": 3.795e-06, + "loss": 0.0926, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.46945369243621826, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1016, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.6481541395187378, + "learning_rate": 3.895e-06, + "loss": 0.0917, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.6769195795059204, + "learning_rate": 3.945e-06, + "loss": 0.0924, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5225585103034973, + "learning_rate": 3.995e-06, + "loss": 0.0909, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.642867922782898, + "learning_rate": 4.045e-06, + "loss": 0.0829, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.4617709219455719, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0924, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.5350925326347351, + "learning_rate": 4.145e-06, + "loss": 0.0821, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.4871961176395416, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0896, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.5545480847358704, + "learning_rate": 4.245e-06, + "loss": 0.0891, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.6057930588722229, + "learning_rate": 4.295e-06, + "loss": 0.0876, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.515282154083252, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0873, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.5402254462242126, + "learning_rate": 4.395e-06, + "loss": 0.0908, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4669134318828583, + "learning_rate": 4.445e-06, + "loss": 0.0837, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.46819189190864563, + "learning_rate": 4.495e-06, + "loss": 0.0888, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35766035318374634, + "learning_rate": 4.545e-06, + "loss": 0.0829, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.6769008040428162, + "learning_rate": 4.595e-06, + "loss": 0.0817, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5884619355201721, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0892, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5613101720809937, + "learning_rate": 4.695e-06, + "loss": 0.0852, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.47087088227272034, + "learning_rate": 4.745e-06, + "loss": 0.0818, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.4887458384037018, + "learning_rate": 4.795e-06, + "loss": 0.0858, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.44256526231765747, + "learning_rate": 4.845e-06, + "loss": 0.0833, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5375688076019287, + "learning_rate": 4.895e-06, + "loss": 0.075, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.6297019124031067, + "learning_rate": 4.945e-06, + "loss": 0.0932, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.6576125621795654, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0829, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.724773108959198, + "learning_rate": 5.045e-06, + "loss": 0.0813, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.54575115442276, + "learning_rate": 5.095e-06, + "loss": 0.0873, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.5186223387718201, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0736, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.46651381254196167, + "learning_rate": 5.195e-06, + "loss": 0.0791, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.5323564410209656, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0753, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.42579638957977295, + "learning_rate": 5.295e-06, + "loss": 0.0755, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.5063076019287109, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0767, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.4558819532394409, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0807, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.6920697093009949, + "learning_rate": 5.445e-06, + "loss": 0.079, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.4706108868122101, + "learning_rate": 5.495e-06, + "loss": 0.079, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.46705493330955505, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0745, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4643767774105072, + "learning_rate": 5.595e-06, + "loss": 0.0778, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.4800759553909302, + "learning_rate": 5.645e-06, + "loss": 0.0792, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.37705323100090027, + "learning_rate": 5.695e-06, + "loss": 0.0815, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.6651142239570618, + "learning_rate": 5.745e-06, + "loss": 0.0732, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.5275036692619324, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0758, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.45047008991241455, + "learning_rate": 5.845e-06, + "loss": 0.0808, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.6931971311569214, + "learning_rate": 5.895e-06, + "loss": 0.0808, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.7053916454315186, + "learning_rate": 5.945e-06, + "loss": 0.0747, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.6798260807991028, + "learning_rate": 5.995e-06, + "loss": 0.0829, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.5724003314971924, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0749, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.5836488008499146, + "learning_rate": 6.095e-06, + "loss": 0.0759, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.40817850828170776, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0778, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.41070470213890076, + "learning_rate": 6.195e-06, + "loss": 0.0765, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.43642961978912354, + "learning_rate": 6.245e-06, + "loss": 0.0711, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.49526578187942505, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0713, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.7773705720901489, + "learning_rate": 6.345e-06, + "loss": 0.0744, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.5485936999320984, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0732, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.5593521595001221, + "learning_rate": 6.445e-06, + "loss": 0.0747, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.3667270839214325, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0723, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.5555917620658875, + "learning_rate": 6.545e-06, + "loss": 0.0726, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.38107141852378845, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0647, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.46452224254608154, + "learning_rate": 6.645e-06, + "loss": 0.0787, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6018183827400208, + "learning_rate": 6.695e-06, + "loss": 0.0718, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.44997578859329224, + "learning_rate": 6.745e-06, + "loss": 0.0758, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45717573165893555, + "learning_rate": 6.795e-06, + "loss": 0.0738, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.5185182690620422, + "learning_rate": 6.845e-06, + "loss": 0.0768, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.5135819315910339, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.43529245257377625, + "learning_rate": 6.945e-06, + "loss": 0.0691, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.49389752745628357, + "learning_rate": 6.995e-06, + "loss": 0.0678, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.4113156199455261, + "learning_rate": 7.045e-06, + "loss": 0.0651, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.4072611927986145, + "learning_rate": 7.095e-06, + "loss": 0.0691, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.4524228870868683, + "learning_rate": 7.145e-06, + "loss": 0.0688, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.6225605607032776, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0676, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4841763377189636, + "learning_rate": 7.245e-06, + "loss": 0.0678, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.46297937631607056, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0759, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.5585324168205261, + "learning_rate": 7.345e-06, + "loss": 0.0671, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.57328200340271, + "learning_rate": 7.395e-06, + "loss": 0.0717, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.5072467923164368, + "learning_rate": 7.445000000000001e-06, + "loss": 0.071, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.4630240797996521, + "learning_rate": 7.495e-06, + "loss": 0.0674, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.35908272862434387, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0704, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.5117342472076416, + "learning_rate": 7.595e-06, + "loss": 0.0675, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.5214630961418152, + "learning_rate": 7.645e-06, + "loss": 0.0702, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.39036089181900024, + "learning_rate": 7.695e-06, + "loss": 0.0709, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.558722734451294, + "learning_rate": 7.745e-06, + "loss": 0.0705, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.3612176179885864, + "learning_rate": 7.795e-06, + "loss": 0.0715, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.523444652557373, + "learning_rate": 7.845e-06, + "loss": 0.0685, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.5231291651725769, + "learning_rate": 7.895e-06, + "loss": 0.0717, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.50368332862854, + "learning_rate": 7.945e-06, + "loss": 0.0694, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.4126806855201721, + "learning_rate": 7.995e-06, + "loss": 0.0693, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.719352662563324, + "learning_rate": 8.045e-06, + "loss": 0.0651, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.4405677020549774, + "learning_rate": 8.095e-06, + "loss": 0.0645, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.34746286273002625, + "learning_rate": 8.145e-06, + "loss": 0.0687, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.43200597167015076, + "learning_rate": 8.195e-06, + "loss": 0.0655, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.3464098274707794, + "learning_rate": 8.245e-06, + "loss": 0.0639, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.27263233065605164, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0629, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.343181848526001, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3949901759624481, + "learning_rate": 8.395e-06, + "loss": 0.0643, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.4489574730396271, + "learning_rate": 8.445e-06, + "loss": 0.0698, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.4746992588043213, + "learning_rate": 8.495e-06, + "loss": 0.0702, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.3651655316352844, + "learning_rate": 8.545e-06, + "loss": 0.0673, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.471071720123291, + "learning_rate": 8.595e-06, + "loss": 0.0648, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.35623008012771606, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0698, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.41837432980537415, + "learning_rate": 8.695e-06, + "loss": 0.0639, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.45608940720558167, + "learning_rate": 8.745e-06, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.40346869826316833, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0625, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.479824960231781, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0703, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.38570791482925415, + "learning_rate": 8.895e-06, + "loss": 0.0643, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.36530473828315735, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0652, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37868866324424744, + "learning_rate": 8.995000000000001e-06, + "loss": 0.068, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.32546791434288025, + "learning_rate": 9.045e-06, + "loss": 0.0636, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4664970338344574, + "learning_rate": 9.095e-06, + "loss": 0.0671, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4767569601535797, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0625, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.41823142766952515, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0685, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.3709985613822937, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0681, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.3933296799659729, + "learning_rate": 9.295e-06, + "loss": 0.0678, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.3079732060432434, + "learning_rate": 9.345e-06, + "loss": 0.0598, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.33485978841781616, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0644, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.40312913060188293, + "learning_rate": 9.445e-06, + "loss": 0.0684, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.4334392547607422, + "learning_rate": 9.495e-06, + "loss": 0.0622, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.5787541270256042, + "learning_rate": 9.545e-06, + "loss": 0.0656, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.2827294170856476, + "learning_rate": 9.595e-06, + "loss": 0.0628, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4431678056716919, + "learning_rate": 9.645e-06, + "loss": 0.0633, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.3916102647781372, + "learning_rate": 9.695e-06, + "loss": 0.0632, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.41470348834991455, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0612, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.3224067986011505, + "learning_rate": 9.795e-06, + "loss": 0.061, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.4110186696052551, + "learning_rate": 9.845e-06, + "loss": 0.066, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.4786360561847687, + "learning_rate": 9.895e-06, + "loss": 0.0648, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3667179048061371, + "learning_rate": 9.945e-06, + "loss": 0.0663, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.31722185015678406, + "learning_rate": 9.995e-06, + "loss": 0.0588, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.29271745681762695, + "learning_rate": 1.0045e-05, + "loss": 0.0636, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.3536592423915863, + "learning_rate": 1.0095e-05, + "loss": 0.0625, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.3228510916233063, + "learning_rate": 1.0145e-05, + "loss": 0.067, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.38474392890930176, + "learning_rate": 1.0195e-05, + "loss": 0.0609, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.40410760045051575, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0638, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.3632320463657379, + "learning_rate": 1.0295e-05, + "loss": 0.0587, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.37775692343711853, + "learning_rate": 1.0345e-05, + "loss": 0.0673, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.3343006372451782, + "learning_rate": 1.0395e-05, + "loss": 0.0599, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3232511281967163, + "learning_rate": 1.0445e-05, + "loss": 0.0661, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.38730430603027344, + "learning_rate": 1.0495e-05, + "loss": 0.0649, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.38972100615501404, + "learning_rate": 1.0545e-05, + "loss": 0.0582, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.27242428064346313, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0583, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.30514073371887207, + "learning_rate": 1.0645e-05, + "loss": 0.065, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.45364341139793396, + "learning_rate": 1.0695e-05, + "loss": 0.0576, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.3863564729690552, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0579, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.33717092871665955, + "learning_rate": 1.0795e-05, + "loss": 0.0565, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38073790073394775, + "learning_rate": 1.0845e-05, + "loss": 0.0585, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4277617037296295, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0622, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.3545698821544647, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0609, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.30646640062332153, + "learning_rate": 1.0995e-05, + "loss": 0.0659, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.3607352375984192, + "learning_rate": 1.1045e-05, + "loss": 0.0621, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3624425530433655, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0626, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3856692612171173, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0604, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.4390200972557068, + "learning_rate": 1.1195e-05, + "loss": 0.0601, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.4108872711658478, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0618, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.32732832431793213, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0636, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.3602239787578583, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.29978248476982117, + "learning_rate": 1.1395e-05, + "loss": 0.058, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.48490920662879944, + "learning_rate": 1.1445e-05, + "loss": 0.0614, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.3278113901615143, + "learning_rate": 1.1495e-05, + "loss": 0.0611, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.37862712144851685, + "learning_rate": 1.1545e-05, + "loss": 0.0626, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.3514714241027832, + "learning_rate": 1.1595e-05, + "loss": 0.0583, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.27256545424461365, + "learning_rate": 1.1645e-05, + "loss": 0.062, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.3305833339691162, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0572, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.2881031930446625, + "learning_rate": 1.1745e-05, + "loss": 0.0641, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.3219377398490906, + "learning_rate": 1.1795e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.34189504384994507, + "learning_rate": 1.1845e-05, + "loss": 0.0607, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30541297793388367, + "learning_rate": 1.1895e-05, + "loss": 0.056, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.35898879170417786, + "learning_rate": 1.1945e-05, + "loss": 0.0598, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.33221277594566345, + "learning_rate": 1.1995e-05, + "loss": 0.0603, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.3435590863227844, + "learning_rate": 1.2045e-05, + "loss": 0.0553, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.24820426106452942, + "learning_rate": 1.2095e-05, + "loss": 0.0644, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.3749466836452484, + "learning_rate": 1.2145e-05, + "loss": 0.0615, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.3984697759151459, + "learning_rate": 1.2195e-05, + "loss": 0.0595, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3250048756599426, + "learning_rate": 1.2245e-05, + "loss": 0.0608, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.2794302701950073, + "learning_rate": 1.2295e-05, + "loss": 0.0578, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3811834752559662, + "learning_rate": 1.2345e-05, + "loss": 0.0612, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3160156309604645, + "learning_rate": 1.2395e-05, + "loss": 0.0541, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.3825189173221588, + "learning_rate": 1.2445e-05, + "loss": 0.0621, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.3966415226459503, + "learning_rate": 1.2495e-05, + "loss": 0.0598, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3615199625492096, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0616, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.440762996673584, + "learning_rate": 1.2595e-05, + "loss": 0.0599, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.345599889755249, + "learning_rate": 1.2645e-05, + "loss": 0.0597, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.2963278889656067, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0556, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.3720734417438507, + "learning_rate": 1.2745e-05, + "loss": 0.0618, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3937445878982544, + "learning_rate": 1.2795e-05, + "loss": 0.0624, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.4492146372795105, + "learning_rate": 1.2845e-05, + "loss": 0.0573, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.5172158479690552, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0594, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.3796320855617523, + "learning_rate": 1.2945e-05, + "loss": 0.0528, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.42423033714294434, + "learning_rate": 1.2995e-05, + "loss": 0.0615, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.4578545093536377, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0562, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.38345491886138916, + "learning_rate": 1.3095e-05, + "loss": 0.0574, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.2539816200733185, + "learning_rate": 1.3145e-05, + "loss": 0.0562, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.44323891401290894, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0588, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.4074377417564392, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0591, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.5252887606620789, + "learning_rate": 1.3295e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.39536044001579285, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0588, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.6300645470619202, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0584, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.312466561794281, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0588, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3810383379459381, + "learning_rate": 1.3495e-05, + "loss": 0.0597, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.30645400285720825, + "learning_rate": 1.3545e-05, + "loss": 0.0589, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.2857605218887329, + "learning_rate": 1.3595e-05, + "loss": 0.0589, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.3206622898578644, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0597, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3226073086261749, + "learning_rate": 1.3695e-05, + "loss": 0.0559, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.24912619590759277, + "learning_rate": 1.3745e-05, + "loss": 0.0547, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.29700762033462524, + "learning_rate": 1.3795e-05, + "loss": 0.0563, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.32627588510513306, + "learning_rate": 1.3845e-05, + "loss": 0.0602, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.48569780588150024, + "learning_rate": 1.3895e-05, + "loss": 0.0553, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.323607474565506, + "learning_rate": 1.3945e-05, + "loss": 0.0578, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.42626142501831055, + "learning_rate": 1.3995e-05, + "loss": 0.0539, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.3281888961791992, + "learning_rate": 1.4045e-05, + "loss": 0.0638, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.38216397166252136, + "learning_rate": 1.4095e-05, + "loss": 0.0572, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.2816051244735718, + "learning_rate": 1.4145e-05, + "loss": 0.0543, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.2604160010814667, + "learning_rate": 1.4195e-05, + "loss": 0.0555, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.34318771958351135, + "learning_rate": 1.4245e-05, + "loss": 0.0565, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.31248190999031067, + "learning_rate": 1.4295e-05, + "loss": 0.0583, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.3014426827430725, + "learning_rate": 1.4345e-05, + "loss": 0.0586, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3650360703468323, + "learning_rate": 1.4395e-05, + "loss": 0.0581, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.33519116044044495, + "learning_rate": 1.4445e-05, + "loss": 0.0611, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.3595008850097656, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.054, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.32799217104911804, + "learning_rate": 1.4545e-05, + "loss": 0.0574, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.29410457611083984, + "learning_rate": 1.4595e-05, + "loss": 0.0582, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.2727576792240143, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0588, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.36562401056289673, + "learning_rate": 1.4695e-05, + "loss": 0.0597, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.3351168930530548, + "learning_rate": 1.4745e-05, + "loss": 0.0558, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.23880337178707123, + "learning_rate": 1.4795e-05, + "loss": 0.0571, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.3910059928894043, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0573, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.3103587031364441, + "learning_rate": 1.4895e-05, + "loss": 0.0578, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.3265620768070221, + "learning_rate": 1.4945e-05, + "loss": 0.0542, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.3060731291770935, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0537, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.41589322686195374, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0546, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.3102424740791321, + "learning_rate": 1.5095e-05, + "loss": 0.0564, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.32286810874938965, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0519, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.3129369020462036, + "learning_rate": 1.5195e-05, + "loss": 0.0566, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.43735018372535706, + "learning_rate": 1.5245e-05, + "loss": 0.0552, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.440449595451355, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0589, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.3631616532802582, + "learning_rate": 1.5345e-05, + "loss": 0.0511, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.29124221205711365, + "learning_rate": 1.5395e-05, + "loss": 0.055, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.4247622489929199, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0562, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.36976534128189087, + "learning_rate": 1.5495e-05, + "loss": 0.0604, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.3580514192581177, + "learning_rate": 1.5545e-05, + "loss": 0.0547, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.2969493567943573, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0564, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28583595156669617, + "learning_rate": 1.5645e-05, + "loss": 0.0586, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.3220328390598297, + "learning_rate": 1.5695e-05, + "loss": 0.0588, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.32270652055740356, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0559, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.3170712888240814, + "learning_rate": 1.5795e-05, + "loss": 0.0566, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.2966325879096985, + "learning_rate": 1.5845e-05, + "loss": 0.0551, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.2788766622543335, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.051, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.3521708548069, + "learning_rate": 1.5945e-05, + "loss": 0.0558, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.35105040669441223, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0551, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.4221852123737335, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0573, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.4217511713504791, + "learning_rate": 1.6095e-05, + "loss": 0.0532, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.4519175589084625, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0489, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3427059054374695, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0544, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29204127192497253, + "learning_rate": 1.6245e-05, + "loss": 0.0525, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.24138635396957397, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0559, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.3599991500377655, + "learning_rate": 1.6345e-05, + "loss": 0.0544, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.3326258361339569, + "learning_rate": 1.6395e-05, + "loss": 0.0593, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31976261734962463, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0624, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.26379555463790894, + "learning_rate": 1.6495e-05, + "loss": 0.0529, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2576923966407776, + "learning_rate": 1.6545e-05, + "loss": 0.0527, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.30890268087387085, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0571, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.4547998309135437, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.3050350248813629, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0552, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.2984176278114319, + "learning_rate": 1.6745e-05, + "loss": 0.0528, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.34583792090415955, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0512, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.33085328340530396, + "learning_rate": 1.6845e-05, + "loss": 0.0563, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.33391326665878296, + "learning_rate": 1.6895e-05, + "loss": 0.0527, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.23657914996147156, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0509, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.3719659149646759, + "learning_rate": 1.6995e-05, + "loss": 0.0568, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.4268389940261841, + "learning_rate": 1.7045e-05, + "loss": 0.0525, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.2726958394050598, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0482, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27847009897232056, + "learning_rate": 1.7145e-05, + "loss": 0.0559, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.7195e-05, + "loss": 0.0588, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.2666028141975403, + "learning_rate": 1.7245e-05, + "loss": 0.0548, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.26048436760902405, + "learning_rate": 1.7295e-05, + "loss": 0.0592, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.22569318115711212, + "learning_rate": 1.7345e-05, + "loss": 0.0505, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.35350096225738525, + "learning_rate": 1.7395e-05, + "loss": 0.0566, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.3166314363479614, + "learning_rate": 1.7445e-05, + "loss": 0.0517, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.24918541312217712, + "learning_rate": 1.7495e-05, + "loss": 0.0545, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3416207730770111, + "learning_rate": 1.7545e-05, + "loss": 0.0544, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.22633257508277893, + "learning_rate": 1.7595e-05, + "loss": 0.0545, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.3972736597061157, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0591, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3213656544685364, + "learning_rate": 1.7695e-05, + "loss": 0.0537, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.30361926555633545, + "learning_rate": 1.7745e-05, + "loss": 0.0529, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.35521000623703003, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0549, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.4444983899593353, + "learning_rate": 1.7845e-05, + "loss": 0.0529, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.32495826482772827, + "learning_rate": 1.7895e-05, + "loss": 0.0557, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.2893896698951721, + "learning_rate": 1.7945e-05, + "loss": 0.056, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.2607886493206024, + "learning_rate": 1.7995e-05, + "loss": 0.055, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.3305937647819519, + "learning_rate": 1.8045e-05, + "loss": 0.0529, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.3003319799900055, + "learning_rate": 1.8095e-05, + "loss": 0.0515, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3523712456226349, + "learning_rate": 1.8145e-05, + "loss": 0.0524, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.27575406432151794, + "learning_rate": 1.8195e-05, + "loss": 0.0535, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.36445286870002747, + "learning_rate": 1.8245e-05, + "loss": 0.0552, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.41415491700172424, + "learning_rate": 1.8295e-05, + "loss": 0.0537, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.3738953471183777, + "learning_rate": 1.8345e-05, + "loss": 0.0555, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.3446105420589447, + "learning_rate": 1.8395e-05, + "loss": 0.0526, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.39016082882881165, + "learning_rate": 1.8445e-05, + "loss": 0.0536, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.4075721502304077, + "learning_rate": 1.8495e-05, + "loss": 0.0548, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.4071345925331116, + "learning_rate": 1.8545e-05, + "loss": 0.0539, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.34446755051612854, + "learning_rate": 1.8595e-05, + "loss": 0.057, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.3431609272956848, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0515, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.31537994742393494, + "learning_rate": 1.8695e-05, + "loss": 0.0534, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.4846970736980438, + "learning_rate": 1.8745e-05, + "loss": 0.0532, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.3109683394432068, + "learning_rate": 1.8795e-05, + "loss": 0.0547, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.28903284668922424, + "learning_rate": 1.8845e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.26592350006103516, + "learning_rate": 1.8895e-05, + "loss": 0.0499, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.27347302436828613, + "learning_rate": 1.8945e-05, + "loss": 0.052, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.3194405138492584, + "learning_rate": 1.8995e-05, + "loss": 0.0521, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.40457430481910706, + "learning_rate": 1.9045e-05, + "loss": 0.0523, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.3055919408798218, + "learning_rate": 1.9095e-05, + "loss": 0.0504, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.4911873936653137, + "learning_rate": 1.9145e-05, + "loss": 0.0531, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.30619072914123535, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0525, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.24404728412628174, + "learning_rate": 1.9245e-05, + "loss": 0.0498, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.31723177433013916, + "learning_rate": 1.9295e-05, + "loss": 0.0519, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.28105342388153076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0528, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.32965341210365295, + "learning_rate": 1.9395e-05, + "loss": 0.0504, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.2487417608499527, + "learning_rate": 1.9445e-05, + "loss": 0.0552, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3788810670375824, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0535, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.2561226189136505, + "learning_rate": 1.9545e-05, + "loss": 0.056, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.29964715242385864, + "learning_rate": 1.9595e-05, + "loss": 0.0565, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.386600136756897, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0468, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.37690475583076477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.35367539525032043, + "learning_rate": 1.9745e-05, + "loss": 0.0501, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.30189844965934753, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0524, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.27525416016578674, + "learning_rate": 1.9845e-05, + "loss": 0.0524, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.22557462751865387, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.3572654128074646, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0536, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.37130483984947205, + "learning_rate": 1.9995e-05, + "loss": 0.0507, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.40820422768592834, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0547, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.24222403764724731, + "learning_rate": 2.0095e-05, + "loss": 0.0536, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.35473307967185974, + "learning_rate": 2.0145e-05, + "loss": 0.0503, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24238434433937073, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0511, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.32721346616744995, + "learning_rate": 2.0245e-05, + "loss": 0.0474, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.3787407875061035, + "learning_rate": 2.0295e-05, + "loss": 0.0515, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.3435424566268921, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0521, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.250515341758728, + "learning_rate": 2.0395e-05, + "loss": 0.0539, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.2730806767940521, + "learning_rate": 2.0445e-05, + "loss": 0.0544, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.32704228162765503, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0516, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.26865822076797485, + "learning_rate": 2.0545e-05, + "loss": 0.0557, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.31514087319374084, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.051, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.2616795301437378, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0483, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2979176938533783, + "learning_rate": 2.0695e-05, + "loss": 0.056, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.2876877188682556, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0496, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.2801620662212372, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0533, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.3521801233291626, + "learning_rate": 2.0845e-05, + "loss": 0.051, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.3016152083873749, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0515, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.29837727546691895, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0483, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.255982369184494, + "learning_rate": 2.0995e-05, + "loss": 0.0482, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.20852255821228027, + "learning_rate": 2.1045e-05, + "loss": 0.05, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.2746782898902893, + "learning_rate": 2.1095e-05, + "loss": 0.0513, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.28745001554489136, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0486, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.3163246810436249, + "learning_rate": 2.1195e-05, + "loss": 0.0536, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.2279432713985443, + "learning_rate": 2.1245e-05, + "loss": 0.0497, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.2273164689540863, + "learning_rate": 2.1295e-05, + "loss": 0.0485, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.3124007284641266, + "learning_rate": 2.1345e-05, + "loss": 0.0473, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.2097717970609665, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0514, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.24428582191467285, + "learning_rate": 2.1445e-05, + "loss": 0.0543, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.47236987948417664, + "learning_rate": 2.1495e-05, + "loss": 0.0525, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.3367164433002472, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0525, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.34086376428604126, + "learning_rate": 2.1595e-05, + "loss": 0.049, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.3060646951198578, + "learning_rate": 2.1645e-05, + "loss": 0.0538, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.2997030019760132, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0553, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.2652549743652344, + "learning_rate": 2.1745e-05, + "loss": 0.051, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.3072361350059509, + "learning_rate": 2.1795e-05, + "loss": 0.0516, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2494215965270996, + "learning_rate": 2.1845e-05, + "loss": 0.0476, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.20796196162700653, + "learning_rate": 2.1895e-05, + "loss": 0.0553, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.24235095083713531, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.21033242344856262, + "learning_rate": 2.1995e-05, + "loss": 0.0542, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.26056164503097534, + "learning_rate": 2.2045e-05, + "loss": 0.0546, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.24798662960529327, + "learning_rate": 2.2095e-05, + "loss": 0.0529, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.22179707884788513, + "learning_rate": 2.2145e-05, + "loss": 0.0542, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.2660025358200073, + "learning_rate": 2.2195e-05, + "loss": 0.0496, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.3771500289440155, + "learning_rate": 2.2245e-05, + "loss": 0.0502, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.3059975206851959, + "learning_rate": 2.2295e-05, + "loss": 0.0482, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.2941029667854309, + "learning_rate": 2.2345e-05, + "loss": 0.0483, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3481833040714264, + "learning_rate": 2.2395e-05, + "loss": 0.0541, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.2729704678058624, + "learning_rate": 2.2445e-05, + "loss": 0.0514, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.44202038645744324, + "learning_rate": 2.2495e-05, + "loss": 0.0534, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.311088889837265, + "learning_rate": 2.2545e-05, + "loss": 0.0474, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30663055181503296, + "learning_rate": 2.2595e-05, + "loss": 0.0491, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.3530942499637604, + "learning_rate": 2.2645e-05, + "loss": 0.0503, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.27998727560043335, + "learning_rate": 2.2695e-05, + "loss": 0.0493, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.2679221034049988, + "learning_rate": 2.2745e-05, + "loss": 0.0512, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.348821222782135, + "learning_rate": 2.2795e-05, + "loss": 0.0489, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.260958194732666, + "learning_rate": 2.2845e-05, + "loss": 0.0474, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.2429468184709549, + "learning_rate": 2.2895e-05, + "loss": 0.053, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.2635575234889984, + "learning_rate": 2.2945e-05, + "loss": 0.053, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.3416514992713928, + "learning_rate": 2.2995e-05, + "loss": 0.0476, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.221172496676445, + "learning_rate": 2.3045e-05, + "loss": 0.0519, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.30541878938674927, + "learning_rate": 2.3095e-05, + "loss": 0.0522, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.25274574756622314, + "learning_rate": 2.3145e-05, + "loss": 0.0499, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.366253525018692, + "learning_rate": 2.3195e-05, + "loss": 0.0463, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.16825024783611298, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0493, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.3158467710018158, + "learning_rate": 2.3295e-05, + "loss": 0.0482, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.2887287139892578, + "learning_rate": 2.3345e-05, + "loss": 0.053, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.4717058539390564, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0515, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.3148113489151001, + "learning_rate": 2.3445e-05, + "loss": 0.0523, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.23815515637397766, + "learning_rate": 2.3495e-05, + "loss": 0.0523, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23897255957126617, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0484, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.2680213749408722, + "learning_rate": 2.3595e-05, + "loss": 0.0485, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.29999756813049316, + "learning_rate": 2.3645e-05, + "loss": 0.0506, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.30607151985168457, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.05, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.26077136397361755, + "learning_rate": 2.3745e-05, + "loss": 0.0558, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2638765275478363, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.049, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.24569877982139587, + "learning_rate": 2.3845e-05, + "loss": 0.053, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.29622989892959595, + "learning_rate": 2.3895e-05, + "loss": 0.0517, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.2598831355571747, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.05, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.25716227293014526, + "learning_rate": 2.3995e-05, + "loss": 0.0453, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.37305381894111633, + "learning_rate": 2.4045e-05, + "loss": 0.0539, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.3151630461215973, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0516, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.296941339969635, + "learning_rate": 2.4145e-05, + "loss": 0.0509, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.31995704770088196, + "learning_rate": 2.4195e-05, + "loss": 0.0498, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2426208257675171, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0475, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.23927678167819977, + "learning_rate": 2.4295e-05, + "loss": 0.05, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.1887904703617096, + "learning_rate": 2.4345e-05, + "loss": 0.05, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.18480589985847473, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0496, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.24515867233276367, + "learning_rate": 2.4445e-05, + "loss": 0.0464, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.25505101680755615, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0481, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.24435125291347504, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0502, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.25899314880371094, + "learning_rate": 2.4595e-05, + "loss": 0.0535, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.30518922209739685, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0481, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.26098036766052246, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0477, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.24723340570926666, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.26713746786117554, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0506, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.21166373789310455, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0502, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2602159082889557, + "learning_rate": 2.4895e-05, + "loss": 0.0506, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27918240427970886, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.049, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.2131885290145874, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0477, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.28778040409088135, + "learning_rate": 2.5045e-05, + "loss": 0.0487, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.3056010603904724, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.049, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.2850024700164795, + "learning_rate": 2.5145e-05, + "loss": 0.0475, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.20490600168704987, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0497, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.32371339201927185, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0491, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.32466208934783936, + "learning_rate": 2.5295e-05, + "loss": 0.0485, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.2779853343963623, + "learning_rate": 2.5345e-05, + "loss": 0.0485, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.3257012367248535, + "learning_rate": 2.5395e-05, + "loss": 0.0486, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2625649571418762, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0486, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.19846542179584503, + "learning_rate": 2.5495e-05, + "loss": 0.0494, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.2415613830089569, + "learning_rate": 2.5545e-05, + "loss": 0.0539, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.20438073575496674, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0476, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.2543872892856598, + "learning_rate": 2.5645e-05, + "loss": 0.0489, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.26969635486602783, + "learning_rate": 2.5695e-05, + "loss": 0.0557, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.27818888425827026, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0517, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.28826993703842163, + "learning_rate": 2.5795e-05, + "loss": 0.0499, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.19518250226974487, + "learning_rate": 2.5845e-05, + "loss": 0.0499, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.2639068067073822, + "learning_rate": 2.5895e-05, + "loss": 0.0459, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.26014742255210876, + "learning_rate": 2.5945e-05, + "loss": 0.0522, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.26659756898880005, + "learning_rate": 2.5995e-05, + "loss": 0.0471, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.24073593318462372, + "learning_rate": 2.6045e-05, + "loss": 0.0488, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.2724616527557373, + "learning_rate": 2.6095e-05, + "loss": 0.0493, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.3012109398841858, + "learning_rate": 2.6145e-05, + "loss": 0.049, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23306235671043396, + "learning_rate": 2.6195e-05, + "loss": 0.0474, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.23879121243953705, + "learning_rate": 2.6245e-05, + "loss": 0.048, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20701922476291656, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0527, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.28734415769577026, + "learning_rate": 2.6345e-05, + "loss": 0.0469, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.26227426528930664, + "learning_rate": 2.6395e-05, + "loss": 0.0472, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.24196277558803558, + "learning_rate": 2.6445e-05, + "loss": 0.0462, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.3405964970588684, + "learning_rate": 2.6495e-05, + "loss": 0.0479, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.24121852219104767, + "learning_rate": 2.6545e-05, + "loss": 0.0444, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.284856379032135, + "learning_rate": 2.6595e-05, + "loss": 0.0492, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.2515311539173126, + "learning_rate": 2.6645e-05, + "loss": 0.0453, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.29297393560409546, + "learning_rate": 2.6695e-05, + "loss": 0.049, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.2528100311756134, + "learning_rate": 2.6745e-05, + "loss": 0.0477, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.29126009345054626, + "learning_rate": 2.6795e-05, + "loss": 0.0484, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.26489511132240295, + "learning_rate": 2.6845e-05, + "loss": 0.0476, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.33150023221969604, + "learning_rate": 2.6895e-05, + "loss": 0.0475, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.25904157757759094, + "learning_rate": 2.6945e-05, + "loss": 0.0495, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.24830900132656097, + "learning_rate": 2.6995e-05, + "loss": 0.0476, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.277997225522995, + "learning_rate": 2.7045e-05, + "loss": 0.0478, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.3100328743457794, + "learning_rate": 2.7095e-05, + "loss": 0.0501, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.2957077622413635, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0491, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.19402435421943665, + "learning_rate": 2.7195e-05, + "loss": 0.0426, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.20080482959747314, + "learning_rate": 2.7245e-05, + "loss": 0.0464, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.21148671209812164, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0457, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.2299162596464157, + "learning_rate": 2.7345e-05, + "loss": 0.0467, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2296467125415802, + "learning_rate": 2.7395e-05, + "loss": 0.0489, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.25099772214889526, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0487, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2503034472465515, + "learning_rate": 2.7495e-05, + "loss": 0.0533, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.24678738415241241, + "learning_rate": 2.7545e-05, + "loss": 0.0462, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.3281477987766266, + "learning_rate": 2.7595e-05, + "loss": 0.0492, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.32214412093162537, + "learning_rate": 2.7645e-05, + "loss": 0.0459, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.29244086146354675, + "learning_rate": 2.7695e-05, + "loss": 0.0495, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.2372455894947052, + "learning_rate": 2.7745e-05, + "loss": 0.0477, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.28411489725112915, + "learning_rate": 2.7795e-05, + "loss": 0.0514, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.36217001080513, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2913915514945984, + "learning_rate": 2.7895e-05, + "loss": 0.0535, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.3237503170967102, + "learning_rate": 2.7945e-05, + "loss": 0.0509, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.32016751170158386, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0526, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.25484171509742737, + "learning_rate": 2.8045e-05, + "loss": 0.0508, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.22641469538211823, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.23269280791282654, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0484, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23346202075481415, + "learning_rate": 2.8195e-05, + "loss": 0.0472, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.25522616505622864, + "learning_rate": 2.8245e-05, + "loss": 0.0493, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.2478165179491043, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0468, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.25732970237731934, + "learning_rate": 2.8345e-05, + "loss": 0.0484, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.21047812700271606, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0455, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.252676397562027, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0496, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20048962533473969, + "learning_rate": 2.8495e-05, + "loss": 0.0434, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.1821785420179367, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0502, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.24106337130069733, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0471, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.413974404335022, + "learning_rate": 2.8645e-05, + "loss": 0.0542, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.29419073462486267, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0499, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.20694036781787872, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0453, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2390667200088501, + "learning_rate": 2.8795e-05, + "loss": 0.045, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2563699781894684, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.046, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.22678422927856445, + "learning_rate": 2.8895e-05, + "loss": 0.0456, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.26104822754859924, + "learning_rate": 2.8945e-05, + "loss": 0.0446, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.22175569832324982, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0478, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.20990405976772308, + "learning_rate": 2.9045e-05, + "loss": 0.0459, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.27277135848999023, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0477, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.22422750294208527, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.047, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.2740926742553711, + "learning_rate": 2.9195e-05, + "loss": 0.0444, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2977927327156067, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0477, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26431694626808167, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0488, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.22759316861629486, + "learning_rate": 2.9345e-05, + "loss": 0.0465, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.24671563506126404, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0501, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.24065586924552917, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0439, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.3334764838218689, + "learning_rate": 2.9495e-05, + "loss": 0.0475, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.23565605282783508, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.046, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.26102808117866516, + "learning_rate": 2.9595e-05, + "loss": 0.0494, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.34544840455055237, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.05, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.2476256936788559, + "learning_rate": 2.9695e-05, + "loss": 0.0477, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.2795628309249878, + "learning_rate": 2.9745e-05, + "loss": 0.0508, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.2756715714931488, + "learning_rate": 2.9795e-05, + "loss": 0.0484, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.20578329265117645, + "learning_rate": 2.9845e-05, + "loss": 0.0461, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.20354008674621582, + "learning_rate": 2.9895e-05, + "loss": 0.0431, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24710936844348907, + "learning_rate": 2.9945e-05, + "loss": 0.0459, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.32639119029045105, + "learning_rate": 2.9995e-05, + "loss": 0.045, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.27762553095817566, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0476, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.2903423607349396, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.3292105793952942, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0503, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.23503656685352325, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0446, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.24153585731983185, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.048, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.24157844483852386, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.046, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.23378467559814453, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0478, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.20753100514411926, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0527, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.2574090361595154, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0468, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.23341813683509827, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0445, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21378666162490845, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0487, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.2410328984260559, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0488, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.22005778551101685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0478, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.3197212219238281, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.046, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.253745436668396, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0437, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.1991635262966156, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0474, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.20773734152317047, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0439, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.3026197552680969, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0462, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.22027666866779327, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0443, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.2512618899345398, + "learning_rate": 2.99997744429684e-05, + "loss": 0.044, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.23001381754875183, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0479, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.20057332515716553, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0433, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.2831740975379944, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0462, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.18362601101398468, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0472, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.29741403460502625, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0443, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.20039376616477966, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0457, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.14999084174633026, + "learning_rate": 2.999958785135414e-05, + "loss": 0.043, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.18134453892707825, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0416, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.25010257959365845, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0494, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.31700536608695984, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0455, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.25080761313438416, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0436, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.2008715718984604, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0459, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.21441133320331573, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0492, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.23232029378414154, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0459, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.30173608660697937, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0505, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.21179363131523132, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0438, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.21905283629894257, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0439, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.17932383716106415, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0452, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.21132917702198029, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0454, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.22732721269130707, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.047, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.28214913606643677, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0446, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.23979637026786804, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0454, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.3095552921295166, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0449, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.2585015594959259, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0473, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.18226519227027893, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0472, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2568792402744293, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0443, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.27914026379585266, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0423, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.21452564001083374, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0457, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.22575126588344574, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0458, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.24053922295570374, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0453, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.21956735849380493, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.048, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2325354665517807, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0439, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.1692194789648056, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.044, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.25796884298324585, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0416, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.2719244956970215, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0445, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.23312833905220032, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0488, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.24458816647529602, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0473, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.24380528926849365, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0411, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.1836114227771759, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0433, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.21051307022571564, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0462, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.18093563616275787, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.045, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2444952130317688, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0485, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.19429947435855865, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0487, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.2294744998216629, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0479, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.2147439569234848, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0492, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.2777695059776306, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0419, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22944948077201843, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0458, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.2822294235229492, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0429, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.20201396942138672, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0435, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.1961182802915573, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0441, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.1653209924697876, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0473, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.2082156240940094, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0508, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.199684277176857, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0466, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.20871016383171082, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0436, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.22781118750572205, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0504, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2652275562286377, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0467, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.20812813937664032, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0456, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2533996105194092, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0458, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.18139654397964478, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0441, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.27243369817733765, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0472, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.326709121465683, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0487, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.26698657870292664, + "learning_rate": 2.999617967180403e-05, + "loss": 0.05, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.4119804799556732, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0455, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.21175070106983185, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0462, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.22503416240215302, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.0468, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.254734605550766, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0471, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26713427901268005, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0439, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.2739362120628357, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0455, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.26221534609794617, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0451, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26752719283103943, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.23052464425563812, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0481, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.162479966878891, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0435, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.24716691672801971, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0459, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.14981329441070557, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0407, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.23326952755451202, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0459, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.2200203239917755, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.046, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.21501530706882477, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0433, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.22950918972492218, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0454, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.2042679488658905, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0444, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.16468611359596252, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0434, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.2180141657590866, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0447, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.1658746749162674, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.24921023845672607, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0398, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0445, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.25189560651779175, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0429, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.2614912688732147, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0427, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.2890806794166565, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0458, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.24215207993984222, + "learning_rate": 2.999336925698801e-05, + "loss": 0.04, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.1806204915046692, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0435, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.18884161114692688, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0422, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.2760990262031555, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0427, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.22660709917545319, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0459, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.24862384796142578, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.045, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.1911306381225586, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0433, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.2612020969390869, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0414, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.20945195853710175, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.043, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.2436099350452423, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0472, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.23788926005363464, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0455, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.20128341019153595, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0456, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.24002164602279663, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0414, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.2653037905693054, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.042, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.15122903883457184, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0395, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.19299687445163727, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0452, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.2534617781639099, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0445, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.29057401418685913, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0398, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.19744732975959778, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.041, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.25267109274864197, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0439, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2406993955373764, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0445, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.2561663091182709, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0465, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.2137032002210617, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0407, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.3002893030643463, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0433, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.32190272212028503, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0444, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.2076684534549713, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0432, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.31036099791526794, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0431, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.21386218070983887, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0442, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.19012001156806946, + "learning_rate": 2.998948184763285e-05, + "loss": 0.048, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.2310018688440323, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0466, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.1984279453754425, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0443, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.24101053178310394, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0446, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.17791084945201874, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0403, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.2016495168209076, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0404, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.19435155391693115, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.21197110414505005, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0398, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.22241441905498505, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.20825718343257904, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0434, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.24533888697624207, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.046, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.2455352544784546, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0412, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.18937505781650543, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0425, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.22356615960597992, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0418, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.23797748982906342, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0457, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.23130296170711517, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0427, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.20548853278160095, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0439, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19323743879795074, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0432, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.24339382350444794, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0412, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.18780124187469482, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0416, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.19353246688842773, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0461, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.21522872149944305, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0412, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.26847484707832336, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.3781932294368744, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0401, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.3219687342643738, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0423, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.1699237823486328, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0421, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.2174658179283142, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0398, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.2433294802904129, + "learning_rate": 2.998488804893431e-05, + "loss": 0.042, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.2116415947675705, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0405, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.22813065350055695, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0432, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.24252736568450928, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.048, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.2009168565273285, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0431, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.19379755854606628, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.23313777148723602, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0384, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.22056345641613007, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0459, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.3403800129890442, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0439, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.17649835348129272, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0419, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.20594412088394165, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0418, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.2901265323162079, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0405, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.22897225618362427, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0424, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.20961393415927887, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0438, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.28585514426231384, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0477, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2239968478679657, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0398, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.1651894897222519, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0405, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.1731943041086197, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0407, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.1708967089653015, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0417, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.17396977543830872, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0359, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.21327704191207886, + "learning_rate": 2.998095036004704e-05, + "loss": 0.045, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.15585461258888245, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0403, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.21774765849113464, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0403, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.1889292299747467, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0408, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.1639491617679596, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0425, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.23745876550674438, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.043, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19481611251831055, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0453, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.21001650393009186, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0429, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.2114192098379135, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0419, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.2506379783153534, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0392, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.21439281105995178, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0422, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.24763159453868866, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0433, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.1719428300857544, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0422, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.25943970680236816, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0425, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22226977348327637, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0448, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.2018195390701294, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0408, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.20869500935077667, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0443, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.23311494290828705, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.042, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.2501373589038849, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0382, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.2650911509990692, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0377, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.21386376023292542, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0405, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.27908873558044434, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0431, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.18731699883937836, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0402, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.2566622197628021, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0419, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.2418505847454071, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0406, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.199744313955307, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0424, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.20134931802749634, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.042, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.15607263147830963, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0453, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.22614715993404388, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0413, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.22810052335262299, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0433, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.16557516157627106, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0424, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.20254060626029968, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0394, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2068638950586319, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0408, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.27261367440223694, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0433, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.23216570913791656, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0445, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.2546656131744385, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0381, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.21829853951931, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0472, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.17109332978725433, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0396, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.1746513843536377, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0413, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.22411789000034332, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0373, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.21182285249233246, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.038, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.19430142641067505, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0419, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.22196920216083527, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0395, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.20695635676383972, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0407, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.1721169799566269, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0396, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.23676829040050507, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0396, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.3019765317440033, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0443, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.23958736658096313, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0402, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.17959944903850555, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0434, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.1563226878643036, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0397, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.16746941208839417, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0402, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.26673364639282227, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0411, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.23634585738182068, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0427, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.19724313914775848, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0408, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.2072090059518814, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0389, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.18574072420597076, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0427, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.23390066623687744, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0411, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.2137259542942047, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0395, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.21458280086517334, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0422, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.1838904619216919, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.043, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.1834891140460968, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0389, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.2012249082326889, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.27410435676574707, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0473, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.21607685089111328, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0405, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.17525449395179749, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.043, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.23137639462947845, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0414, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.20958152413368225, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0404, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.21453802287578583, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.043, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.2522047758102417, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.04, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.23595386743545532, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0396, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.23375281691551208, + "learning_rate": 2.996358559911323e-05, + "loss": 0.042, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.21628527343273163, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0434, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.2463727742433548, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0391, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.19919127225875854, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.043, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.28369203209877014, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0463, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.30222833156585693, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0397, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.19382458925247192, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0402, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2075910121202469, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0392, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.2503066658973694, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0427, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.18662676215171814, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0397, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.2432677149772644, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0425, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.19900213181972504, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0422, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.12742096185684204, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.20504114031791687, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0404, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.19127227365970612, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0425, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.18699967861175537, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0419, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.20074906945228577, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0381, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.24741096794605255, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0389, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.22948721051216125, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0435, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.17772004008293152, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0412, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.228739395737648, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0406, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.18312685191631317, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0401, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.21340692043304443, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0403, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.17110474407672882, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0388, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.19640640914440155, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.2386539727449417, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.043, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.20800922811031342, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0412, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1889747679233551, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0377, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.1801699548959732, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0402, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.20975525677204132, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0426, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.22626695036888123, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0404, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.2192756086587906, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0438, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.21307197213172913, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0447, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.28529244661331177, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0412, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.26367515325546265, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.044, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.29029133915901184, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.14019586145877838, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0427, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19176369905471802, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0408, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.19753597676753998, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0395, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1900634765625, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0427, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.21698379516601562, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0452, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.21722441911697388, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0417, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.1760135143995285, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0405, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18030588328838348, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0425, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.20307210087776184, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0416, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.19713939726352692, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0401, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.18719582259655, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.19560874998569489, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.04, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1633717119693756, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0424, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.20183435082435608, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0397, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.23784901201725006, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0438, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.17358708381652832, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0393, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18991878628730774, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0387, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.20455898344516754, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.042, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.23264537751674652, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0447, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2586922347545624, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0402, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.20447152853012085, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0406, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.1814388930797577, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0412, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2679784595966339, + "learning_rate": 2.994497917870469e-05, + "loss": 0.039, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.1818714141845703, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0423, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.2153957635164261, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0391, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.1858970820903778, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0394, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.23248426616191864, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.17587599158287048, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0367, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.18720309436321259, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.039, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.17602671682834625, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0393, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.15901489555835724, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0418, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.18207870423793793, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0401, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.22800953686237335, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0416, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.20031091570854187, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0417, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.24054989218711853, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0438, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.201461061835289, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0369, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.15697605907917023, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0397, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.16894802451133728, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0403, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2567543089389801, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0421, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.20297718048095703, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0396, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.17331846058368683, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.038, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.2214743047952652, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0405, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.2107590287923813, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0388, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.1768934726715088, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0364, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.16018562018871307, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0405, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.18402984738349915, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0425, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.21482129395008087, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0399, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.20554907619953156, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0388, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.2401154637336731, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0369, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.15072989463806152, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0359, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.1649719625711441, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0399, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.2386736422777176, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0403, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.2389005720615387, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0416, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.14930105209350586, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0372, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.18874907493591309, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0356, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.21466751396656036, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0391, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1697709560394287, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.24686385691165924, + "learning_rate": 2.993190369330068e-05, + "loss": 0.04, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.19733406603336334, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0417, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.3981468677520752, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0375, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.20926636457443237, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0384, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.18989330530166626, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0365, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17171184718608856, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0386, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.20474061369895935, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0413, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.19314105808734894, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0358, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.1794378012418747, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0409, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.17263059318065643, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0385, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.17863675951957703, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0355, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.20071984827518463, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0386, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.2025180608034134, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0409, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.21840184926986694, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0397, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.197452113032341, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0379, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.15370187163352966, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0379, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.22221392393112183, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0396, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.18536874651908875, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.18744619190692902, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0378, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.21197383105754852, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0399, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.29221004247665405, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0355, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.2411249279975891, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.039, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.3187369108200073, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.036, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.18096214532852173, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0372, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.2743191719055176, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0401, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.1864418387413025, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0384, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.15518276393413544, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.039, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.18089881539344788, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0359, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.2218618243932724, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0427, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.17051172256469727, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0401, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.15884852409362793, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0378, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1561235636472702, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0383, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.19711007177829742, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0409, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.18595683574676514, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.043, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.18613699078559875, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0357, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.1390007734298706, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0356, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.25603482127189636, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0377, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.1902502328157425, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0381, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.18781402707099915, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0393, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.20504912734031677, + "learning_rate": 2.991569760600352e-05, + "loss": 0.042, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.17359918355941772, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0373, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.2029414027929306, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0412, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.23485742509365082, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0398, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1454097479581833, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0394, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.22343166172504425, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0386, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.17754614353179932, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0358, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.14276635646820068, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0405, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.2062550187110901, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0429, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2093406766653061, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.038, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.14510402083396912, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0374, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.22726939618587494, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0373, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.14905233681201935, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0392, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.15828309953212738, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0361, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.990945968675917e-05, + "loss": 0.037, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.1867590844631195, + "learning_rate": 2.990900562792633e-05, + "loss": 0.038, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.18278446793556213, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0358, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.24628379940986633, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0393, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.2195759266614914, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0372, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.2397889941930771, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0352, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.1563243269920349, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0361, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.16278347373008728, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0366, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1776551604270935, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0369, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.19893717765808105, + "learning_rate": 2.990533239951709e-05, + "loss": 0.039, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.21403104066848755, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0378, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.19885864853858948, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0393, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.18425336480140686, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0404, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.1930677592754364, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0369, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18271560966968536, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0351, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.16518554091453552, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0367, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.17990152537822723, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0349, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.19625109434127808, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0368, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.1891077160835266, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0407, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.20162540674209595, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.17586444318294525, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0346, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.13146457076072693, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0338, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.16059468686580658, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0366, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.1724517047405243, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0395, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.1847251057624817, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0365, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.20104378461837769, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.037, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.2712373733520508, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0347, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.18014875054359436, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0388, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.17479945719242096, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0388, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.24176514148712158, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0387, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.2453809529542923, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0381, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2105131894350052, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0369, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1763688176870346, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0384, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.16223029792308807, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0348, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.173785999417305, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.038, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.23269419372081757, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0393, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.19176329672336578, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0399, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.1559731364250183, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0375, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.15701691806316376, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.22872254252433777, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0371, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15535064041614532, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0378, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.21613062918186188, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0382, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.1504523903131485, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0386, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.177171990275383, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2307550013065338, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0354, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.2215310037136078, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0371, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.23132716119289398, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0389, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.1937142014503479, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0373, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.23848143219947815, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0355, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.23107734322547913, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0359, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1455046832561493, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0404, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.21890339255332947, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0421, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.1897345334291458, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0357, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.2016463577747345, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0363, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.15850050747394562, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0391, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.18169008195400238, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0377, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.26095741987228394, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0391, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.20983633399009705, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0355, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.21026623249053955, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.2104070633649826, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0343, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1801508367061615, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0369, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.22245049476623535, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.0358, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.19220919907093048, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0354, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.2285826951265335, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0391, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.15382888913154602, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0374, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.13551367819309235, + "learning_rate": 2.987759237035982e-05, + "loss": 0.039, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.23362696170806885, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0367, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.15972478687763214, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0319, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.23858226835727692, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0354, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.17988596856594086, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0391, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.21262454986572266, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0392, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.1443355828523636, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0384, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.24133095145225525, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0382, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.175334632396698, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.039, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.20340824127197266, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0376, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.16549751162528992, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0371, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.16527467966079712, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0373, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1418737769126892, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0378, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18754976987838745, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0355, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.13649867475032806, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0329, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.21408453583717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0348, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.20929333567619324, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.24083013832569122, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0378, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17924576997756958, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0369, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.22360439598560333, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0382, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.15963956713676453, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0361, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.19003787636756897, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0365, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2458433210849762, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0375, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.22048939764499664, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0382, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.19965334236621857, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0371, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.228081613779068, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0375, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.24077889323234558, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0347, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.17692111432552338, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0365, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.1996857225894928, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0358, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.24799935519695282, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0377, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.24300269782543182, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0362, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.2230309098958969, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0364, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.16640354692935944, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0346, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.14670157432556152, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.039, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18531887233257294, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0353, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.19566190242767334, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0391, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.1967019885778427, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0355, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.2161232978105545, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0375, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.17630350589752197, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0397, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18905608355998993, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0347, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.1890578716993332, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.036, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.15397298336029053, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0401, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0362, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.20983757078647614, + "learning_rate": 2.985388676461079e-05, + "loss": 0.036, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.1554982215166092, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0378, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.1961488276720047, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0356, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.17265456914901733, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0357, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.1275556981563568, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0353, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.1667783409357071, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0354, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.27920690178871155, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0385, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.12451875954866409, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0352, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.15794622898101807, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0339, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.16285695135593414, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0369, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.18990163505077362, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0376, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.1867877095937729, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0363, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.15270273387432098, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0394, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.1934751272201538, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0382, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16943520307540894, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0369, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.20546840131282806, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0336, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.18490619957447052, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0415, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.16439616680145264, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0384, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.13623777031898499, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0379, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.18209043145179749, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0357, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.19922122359275818, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0365, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.15637168288230896, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0323, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1781233251094818, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0329, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.17547845840454102, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0344, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.1770639270544052, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0358, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.2677673101425171, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0356, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.16740617156028748, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0344, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.13076575100421906, + "learning_rate": 2.983793562696423e-05, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.17586535215377808, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0381, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.1623464673757553, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.037, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.16550037264823914, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0373, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.20314498245716095, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0366, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.19387134909629822, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0377, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.17531387507915497, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0381, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.21466659009456635, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.037, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1693795919418335, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0339, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.21929262578487396, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0332, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.2253660261631012, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0339, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.2084304839372635, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0363, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.2513981759548187, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0357, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.19585898518562317, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0353, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.22262930870056152, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0357, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.19813011586666107, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0347, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.16626323759555817, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0346, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.271424800157547, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.036, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2283792346715927, + "learning_rate": 2.982684510938631e-05, + "loss": 0.04, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.15996390581130981, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0394, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.20269536972045898, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0387, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.20575514435768127, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0386, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.15178649127483368, + "learning_rate": 2.982433099619666e-05, + "loss": 0.037, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.23187878727912903, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0355, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.12988153100013733, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0369, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19367633759975433, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0367, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.15319304168224335, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0337, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.19311629235744476, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0383, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.14771778881549835, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0393, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.20381954312324524, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0414, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22292324900627136, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0358, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.13758333027362823, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0359, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.18837112188339233, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0339, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.1757460981607437, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0352, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.24256370961666107, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0346, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.2417164295911789, + "learning_rate": 2.981603574229741e-05, + "loss": 0.037, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.21129392087459564, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0371, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16878797113895416, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0367, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.17526096105575562, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0353, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.1812993884086609, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.037, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.16877815127372742, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0353, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.20823006331920624, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0384, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1536571979522705, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0356, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.1477498710155487, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0346, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.12916848063468933, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0318, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.15418170392513275, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0369, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.1883322149515152, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0369, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.1569332480430603, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0374, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.21140694618225098, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0365, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.26859667897224426, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0395, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.2182101309299469, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0382, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.18896763026714325, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0372, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.17398814857006073, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0338, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.14043423533439636, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0361, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.14995983242988586, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0363, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.15542328357696533, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0361, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.16333894431591034, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0337, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.18489807844161987, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0352, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.1794673502445221, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.037, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.1882123351097107, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0414, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.16483181715011597, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.04, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.1727719008922577, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0361, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.1968923956155777, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0389, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1838955134153366, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.034, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.23556186258792877, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0369, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.2518751323223114, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0331, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18682284653186798, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0371, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.2049458622932434, + "learning_rate": 2.979480675356209e-05, + "loss": 0.04, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.11846950650215149, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0347, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.21428292989730835, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0354, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.154558002948761, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0359, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1178995817899704, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0364, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.16926616430282593, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0373, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.1721247136592865, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0346, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.1975203901529312, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0351, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.17340779304504395, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0347, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.22424420714378357, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0384, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.1688208281993866, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.16542011499404907, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0343, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.15097306668758392, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0372, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.15530438721179962, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0345, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.16840992867946625, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.034, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.19959625601768494, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0375, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.15717248618602753, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0344, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.21473488211631775, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.04, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.18634898960590363, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0346, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.16032545268535614, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0377, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.19153323769569397, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0408, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.16018801927566528, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0334, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.1498652696609497, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0349, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.1652575582265854, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0344, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.15636402368545532, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0331, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.22963644564151764, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0359, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.16931135952472687, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0389, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.17646029591560364, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0322, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.19198739528656006, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0318, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.16826699674129486, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0389, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.1693607121706009, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.037, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.147409588098526, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0343, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.1984914094209671, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0329, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.18371893465518951, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0357, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.16279958188533783, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0377, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2122160941362381, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0367, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.20099543035030365, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0343, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.17249372601509094, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0325, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1106863021850586, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0336, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.2338913530111313, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0331, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.1545400768518448, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0321, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.17114387452602386, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0323, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.16218030452728271, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.037, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.14618389308452606, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0306, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.17804935574531555, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0348, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1512002944946289, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0318, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.14877542853355408, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.037, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.1693819761276245, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0325, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.17713536322116852, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0356, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.1719064563512802, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0358, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.19448526203632355, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0363, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.18047156929969788, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0326, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.20896470546722412, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0362, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.15999653935432434, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0318, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.17288383841514587, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0336, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.21483922004699707, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0367, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.19133612513542175, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0336, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.16010481119155884, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0377, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.1778886914253235, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0354, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.14997199177742004, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0326, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.16331784427165985, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0326, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.18247577548027039, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0367, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1838751882314682, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0338, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.17194764316082, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0365, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.1354181170463562, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0315, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.19473528861999512, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0345, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14786818623542786, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0354, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.1931883543729782, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0345, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.159915491938591, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0357, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14373569190502167, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0309, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.1468040943145752, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0339, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.1592605859041214, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0374, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.17073556780815125, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0365, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.13732466101646423, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0331, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.18718381226062775, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0337, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.179783895611763, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0383, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.15795794129371643, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0368, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.13824236392974854, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0343, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.2081076204776764, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0385, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.1407512128353119, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0338, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.1619846224784851, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0325, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.184702068567276, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0353, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.18360818922519684, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0349, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.17683662474155426, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0332, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.17690211534500122, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0364, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.18250274658203125, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0336, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.12902002036571503, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0344, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.1598006635904312, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0332, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.17691342532634735, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.037, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.2213180959224701, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0324, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.14762577414512634, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0321, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.14543993771076202, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.034, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.18970227241516113, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0377, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.15049827098846436, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0341, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.1485714316368103, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0361, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1893187314271927, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0354, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.2345002293586731, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0336, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.24119973182678223, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0325, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.18577414751052856, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0337, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.1941320151090622, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0355, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.17646151781082153, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0342, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.13801687955856323, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0351, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.15124225616455078, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0337, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19114945828914642, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.033, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.23505449295043945, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0357, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.15116260945796967, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.15656375885009766, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0322, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.18982963263988495, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0368, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.18300476670265198, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0379, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.16485247015953064, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0335, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.19399121403694153, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0353, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.22944431006908417, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0343, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.16755999624729156, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0351, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.15779320895671844, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0318, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.2113935351371765, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0343, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.14698673784732819, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0332, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19958600401878357, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0357, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20250289142131805, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0334, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.1955910474061966, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0381, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.13143742084503174, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0321, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.15840105712413788, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0333, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.20588421821594238, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0347, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.1881185919046402, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0345, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.16714425384998322, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.032, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.16526015102863312, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0316, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.1865321844816208, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0351, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.19202300906181335, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0325, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.15551044046878815, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0339, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.15684190392494202, + "learning_rate": 2.969840338753429e-05, + "loss": 0.034, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.16806578636169434, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0319, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.1554284542798996, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0348, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1836974322795868, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0375, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.20003411173820496, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0364, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.1747797578573227, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0328, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.2204974740743637, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0331, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.17183512449264526, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0381, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.14398148655891418, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.03, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.17428989708423615, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0329, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.13864420354366302, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0319, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.15052364766597748, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0357, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.21513856947422028, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0349, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19239531457424164, + "learning_rate": 2.968758727655112e-05, + "loss": 0.0377, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.22284342348575592, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0358, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.17206035554409027, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0349, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.1991468220949173, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0316, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.13542494177818298, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0329, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.14473386108875275, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0346, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.1972355991601944, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0348, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.17447993159294128, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0337, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.14419333636760712, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0355, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.19937965273857117, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0381, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.18988190591335297, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0331, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.1552465558052063, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0333, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.1769426017999649, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0344, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.17652767896652222, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0354, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1598699688911438, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0354, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.17556554079055786, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0372, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.18424467742443085, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0372, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.1591036170721054, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.038, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.1973133534193039, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0332, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.139867901802063, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0336, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.17536301910877228, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0339, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16841425001621246, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0319, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.24087557196617126, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.1419876366853714, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0341, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0315, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.14541317522525787, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0326, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.18411535024642944, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0382, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1769087016582489, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0355, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.16459472477436066, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0354, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.1770380288362503, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.034, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.16219645738601685, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0316, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.18393918871879578, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0366, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.1436525285243988, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0318, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.18282611668109894, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0376, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.16712592542171478, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0367, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.13408221304416656, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0305, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.17664852738380432, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0312, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.13730520009994507, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0323, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1278112679719925, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0322, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.1742715984582901, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0339, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.15022823214530945, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0348, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.17816156148910522, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0348, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.18016202747821808, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0357, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.21158602833747864, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0331, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1747511923313141, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0301, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.16840165853500366, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0363, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.1507093459367752, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0352, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.2081453651189804, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.035, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.1459163874387741, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0331, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.17468400299549103, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0339, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.15824441611766815, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0309, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.19502659142017365, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0305, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.1502636969089508, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.032, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.13870884478092194, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0356, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1596919149160385, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0347, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.1830880492925644, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0359, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.13567551970481873, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0319, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.12907959520816803, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0327, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.17357082664966583, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.0308, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.20002517104148865, + "learning_rate": 2.963613172841619e-05, + "loss": 0.035, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.13852699100971222, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0353, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.15445692837238312, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.11687827110290527, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0319, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.15454575419425964, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0366, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.1916663944721222, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0327, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.16530223190784454, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0309, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17331530153751373, + "learning_rate": 2.962976980219581e-05, + "loss": 0.032, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.17225238680839539, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0323, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.16390499472618103, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0324, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.16459178924560547, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0315, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.1416798084974289, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0348, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.1546057015657425, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0329, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15206973254680634, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0304, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.18791860342025757, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.03, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.17669418454170227, + "learning_rate": 2.962243236871436e-05, + "loss": 0.034, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.1335674524307251, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0323, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.17903365194797516, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0327, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.17771181464195251, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0341, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1532796174287796, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0328, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.18777944147586823, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0343, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.2553282082080841, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0342, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.24115385115146637, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0347, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.18773671984672546, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0324, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.18168678879737854, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0355, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.14390279352664948, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0323, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.13260799646377563, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0336, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.16469483077526093, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0341, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.1618199497461319, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0305, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.15524879097938538, + "learning_rate": 2.960942086834339e-05, + "loss": 0.034, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.14689622819423676, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0352, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.17949312925338745, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0329, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.2376648485660553, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0326, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.1302678883075714, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0356, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.16316285729408264, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0318, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.14314667880535126, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0347, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.13162434101104736, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0335, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.13411720097064972, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0334, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.20853494107723236, + "learning_rate": 2.960094148810064e-05, + "loss": 0.036, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.15634609758853912, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0314, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.16109681129455566, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0323, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.15969422459602356, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0292, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.13126519322395325, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0338, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.13577662408351898, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0323, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13226218521595, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0307, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.12283602356910706, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0305, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16819486021995544, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0316, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.1443309634923935, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.031, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.17339679598808289, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0335, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.12366446852684021, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0319, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.18073806166648865, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0355, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.16230227053165436, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.033, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.15816937386989594, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0331, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.15911494195461273, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0296, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.19500406086444855, + "learning_rate": 2.958564529914256e-05, + "loss": 0.034, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.16553588211536407, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0348, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.1780945360660553, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0297, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16839291155338287, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0317, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.26763999462127686, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0335, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.1292608082294464, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0319, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19598962366580963, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0346, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.16073159873485565, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0317, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.2037622481584549, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0324, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.11946277320384979, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0329, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.12799547612667084, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0329, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.23773977160453796, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0327, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.19241386651992798, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0371, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.15968935191631317, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0335, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.1712830811738968, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0302, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.16464142501354218, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0346, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.2122460901737213, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0319, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.16571341454982758, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.033, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1610543429851532, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0347, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14235250651836395, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0322, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.14999081194400787, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0299, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.033, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.1448807269334793, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0307, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.20123368501663208, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0315, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.16191548109054565, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.031, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.1475522518157959, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.17617955803871155, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0323, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.20633190870285034, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0287, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.13562145829200745, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0347, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.15800844132900238, + "learning_rate": 2.955719840324779e-05, + "loss": 0.034, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.1652759164571762, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0291, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.11819476634263992, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0309, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.17562547326087952, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0327, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.1934351772069931, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0342, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.1565367877483368, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.033, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.21048930287361145, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0357, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.16762706637382507, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0341, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.1267809122800827, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0322, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.18461033701896667, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0357, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.13451576232910156, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0327, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.15478897094726562, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0311, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1461619883775711, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0314, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.1432616114616394, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0339, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.14061163365840912, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0337, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.1950581669807434, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0315, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.13942912220954895, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.031, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.15901322662830353, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0324, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.19862298667430878, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0322, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.15155215561389923, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0302, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16394968330860138, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0299, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.13225089013576508, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0305, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.18203529715538025, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0319, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.15323573350906372, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0346, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.18822363018989563, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.038, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1510896533727646, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0336, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.12274178862571716, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.032, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.1287696361541748, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0341, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.16619601845741272, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0321, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.12426833063364029, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0298, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.1751766800880432, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0341, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.13578996062278748, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0306, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.19475996494293213, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0314, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.13291390240192413, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0341, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.16193236410617828, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0325, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.16108311712741852, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0356, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.16233862936496735, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0303, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.14042457938194275, + "learning_rate": 2.951955450025455e-05, + "loss": 0.031, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.14922280609607697, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0336, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.1389254927635193, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0324, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.21042059361934662, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0331, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.18529878556728363, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0311, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1345708668231964, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0316, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.14794935286045074, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0367, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.18609392642974854, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0326, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.1762595772743225, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0312, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.15549533069133759, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0342, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.1910371482372284, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0316, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.12789981067180634, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0286, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.11457879841327667, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0303, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.12534750998020172, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0313, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.17089471220970154, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0318, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.16774669289588928, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0326, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.18693485856056213, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0356, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.20419354736804962, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15851624310016632, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0328, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1284184604883194, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.1148369237780571, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0291, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.15302640199661255, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.037, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.16082480549812317, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.031, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.16006684303283691, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0346, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14866025745868683, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0323, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.15490789711475372, + "learning_rate": 2.949326462724779e-05, + "loss": 0.033, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.18960629403591156, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0302, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.16329076886177063, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0338, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1700190156698227, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0326, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.14541998505592346, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0332, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.13809162378311157, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.16794940829277039, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0345, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1897762268781662, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0304, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2064688503742218, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0332, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.14193041622638702, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0332, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.15571711957454681, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0339, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.14264710247516632, + "learning_rate": 2.948147907679037e-05, + "loss": 0.036, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.1411963850259781, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0344, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.18891720473766327, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0308, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.15993206202983856, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0339, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.15609434247016907, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0315, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.19272638857364655, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0325, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.2836095690727234, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0323, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16382543742656708, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0335, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.18899215757846832, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0301, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.17764721810817719, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0322, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.1808568239212036, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0304, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.166172593832016, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0298, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.18262767791748047, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0305, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.17854458093643188, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.033, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.16440987586975098, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.033, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.19478020071983337, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0288, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.2605016231536865, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0318, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1850014477968216, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0313, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.1339086890220642, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.17828164994716644, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0315, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.1235261932015419, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0328, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.13704872131347656, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0322, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1296597719192505, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0315, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.16906750202178955, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0347, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14643864333629608, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.031, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.16102853417396545, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0315, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.17575663328170776, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0312, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1488381028175354, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0318, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.14629733562469482, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0333, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.17051763832569122, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.034, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.14636048674583435, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0315, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.19957765936851501, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.032, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.15027561783790588, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0342, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1580146700143814, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.032, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1452188491821289, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0302, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.1924576759338379, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0323, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.199595108628273, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0326, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.1557701975107193, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0303, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.21232250332832336, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0324, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.13840343058109283, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0321, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.14219941198825836, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0314, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1747855842113495, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0342, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1730964183807373, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0334, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.12938249111175537, + "learning_rate": 2.94341324778745e-05, + "loss": 0.029, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1782408058643341, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0299, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.943188093053675e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.12131186574697495, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0277, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.19141671061515808, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0307, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.2431664764881134, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0326, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15145368874073029, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0329, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.19781085848808289, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0332, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.19477561116218567, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0317, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.1818031370639801, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0352, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.13725785911083221, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0272, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.1874770075082779, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0322, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.18701452016830444, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0304, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.15173715353012085, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.032, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.1315767467021942, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0331, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.16627946496009827, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0299, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.2122701108455658, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0308, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.12093837559223175, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0303, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.2072226107120514, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0331, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.14395366609096527, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0323, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.18454021215438843, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0349, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1647966057062149, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0342, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1907777190208435, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0347, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.14438189566135406, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0295, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.16802345216274261, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0283, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.1488419622182846, + "learning_rate": 2.940567320758981e-05, + "loss": 0.033, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14881044626235962, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0318, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.18762606382369995, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0322, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.15426765382289886, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0314, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.17882433533668518, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0361, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.16833186149597168, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0319, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.16899487376213074, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0283, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.15301142632961273, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0325, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.13270439207553864, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0305, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.16603541374206543, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0347, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.21959784626960754, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0299, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.24568438529968262, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0339, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.14889934659004211, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0284, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.20354624092578888, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0309, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.18124797940254211, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0352, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.13871392607688904, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0311, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.18997471034526825, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0299, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.09903991222381592, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0304, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.13336266577243805, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0306, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.1564595103263855, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0295, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.17495550215244293, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.17048844695091248, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0266, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.1257491409778595, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0299, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.15244479477405548, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0341, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.15737923979759216, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.03, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.13226579129695892, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0268, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.21549297869205475, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0323, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.2102872133255005, + "learning_rate": 2.937416924082395e-05, + "loss": 0.029, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.14369380474090576, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0306, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.1390914022922516, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0273, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.13536471128463745, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0299, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.13321688771247864, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0298, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13358351588249207, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0309, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.1732599437236786, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0319, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.17455634474754333, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0314, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.1714484691619873, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0345, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.18120428919792175, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0293, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.21681863069534302, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0335, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.12099924683570862, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0299, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.13233523070812225, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0372, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.16708864271640778, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0305, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.2122814804315567, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0319, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14159013330936432, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0343, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.13396364450454712, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0308, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.17630934715270996, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0349, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.16417403519153595, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0316, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.16593796014785767, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0291, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.13613900542259216, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0299, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.15980054438114166, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0319, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1578991562128067, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0288, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.1628173291683197, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0317, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.15937024354934692, + "learning_rate": 2.934549758921277e-05, + "loss": 0.031, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.15258124470710754, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0323, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.16142596304416656, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0289, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.15643659234046936, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0295, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.17128989100456238, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0313, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.1871197372674942, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0314, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.1448175013065338, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0318, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.21338209509849548, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0334, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15262271463871002, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0293, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.14302000403404236, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0307, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.13756762444972992, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0288, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.1390128880739212, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0286, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.16088609397411346, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0299, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.11852050572633743, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0284, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.17093200981616974, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0331, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.2325763702392578, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0269, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.19948336482048035, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0314, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.16918544471263885, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0311, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.14563684165477753, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0301, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.19645430147647858, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0312, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.18950732052326202, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0292, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.11659497767686844, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0307, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1314595341682434, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0326, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1457567662000656, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.16525784134864807, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0322, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.15977199375629425, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0334, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.17081087827682495, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0308, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.20844970643520355, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.029, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.22166086733341217, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1766587346792221, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0298, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1307239532470703, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0298, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0289, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.17264464497566223, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0317, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.15076857805252075, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0308, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14443998038768768, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0291, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.13115015625953674, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0329, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1444655954837799, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0294, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.16740760207176208, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0318, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.14201150834560394, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0314, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.14225733280181885, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0297, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.16758595407009125, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0307, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.16751281917095184, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.03, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.1855877935886383, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0333, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.17420977354049683, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0277, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.1568649560213089, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0312, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.15874658524990082, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0297, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.13727585971355438, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0334, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.14999760687351227, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0328, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.16789689660072327, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.14890174567699432, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0312, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.18001733720302582, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0322, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.1718987226486206, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.03, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.18727312982082367, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0297, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.1528918594121933, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0296, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.13372628390789032, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0288, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.14980314671993256, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0319, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.11368083208799362, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0312, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.14968374371528625, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0311, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.15593667328357697, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0287, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.1762111783027649, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0296, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.15493011474609375, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0272, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.14142747223377228, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.03, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.1307222545146942, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0294, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.1598598062992096, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.03, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.17966923117637634, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0337, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15533174574375153, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0276, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.1223452091217041, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0294, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.12733329832553864, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0289, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.15903618931770325, + "learning_rate": 2.926085774098786e-05, + "loss": 0.029, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1539735347032547, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0272, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.20487286150455475, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0312, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.19920656085014343, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0315, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.15179695188999176, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0305, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.16460831463336945, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0315, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.21031928062438965, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0283, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.13835182785987854, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0287, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.1597253829240799, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0336, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.11488606780767441, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0328, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.18707934021949768, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0328, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.18874959647655487, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0317, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.1182878389954567, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0316, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.11695742607116699, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0308, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.1597018986940384, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0324, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.17729975283145905, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0316, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.18250048160552979, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0309, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.20568659901618958, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0293, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.12907297909259796, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0281, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.1809365600347519, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0333, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.17379151284694672, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0327, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.16082972288131714, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0308, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.17415443062782288, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0299, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.1659587323665619, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0272, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.1538572907447815, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0309, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.16303570568561554, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0322, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1444631963968277, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0306, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.24458004534244537, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0285, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.15384158492088318, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0277, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1425933837890625, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0303, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.12485234439373016, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0279, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.14287105202674866, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0297, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.1621903032064438, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0268, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.145540252327919, + "learning_rate": 2.921797581830774e-05, + "loss": 0.03, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.1410767287015915, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0309, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.10953503102064133, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0299, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.15029482543468475, + "learning_rate": 2.92140191009304e-05, + "loss": 0.031, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.13262124359607697, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0299, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1503380835056305, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0329, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.18932904303073883, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.028, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.1473148614168167, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0315, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.18515919148921967, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0309, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.16122235357761383, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0279, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.1684419959783554, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0309, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.21400189399719238, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0304, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.12552157044410706, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0286, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.15383073687553406, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.029, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.1269458532333374, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0289, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.1839684545993805, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0316, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.13441112637519836, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0281, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.13045959174633026, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0296, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.13933542370796204, + "learning_rate": 2.919408984114327e-05, + "loss": 0.029, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.15143601596355438, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0296, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.16731251776218414, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.032, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.13245129585266113, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0268, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.17244738340377808, + "learning_rate": 2.918873439576597e-05, + "loss": 0.028, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.1709858626127243, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0293, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18727441132068634, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0302, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1482084095478058, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0297, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.1343366652727127, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0303, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.1456393450498581, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0299, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1569104641675949, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0281, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.19711613655090332, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0298, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.14027468860149384, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.031, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.16852174699306488, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0289, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.18446287512779236, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0308, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.16207733750343323, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0283, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.19747264683246613, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0301, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.15206509828567505, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0295, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.18873688578605652, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0286, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.144051656126976, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0335, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.14620175957679749, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0284, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.19520114362239838, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0304, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.16127106547355652, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0333, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.11467311531305313, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0303, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11777719110250473, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0311, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.12151405215263367, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0278, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1852089911699295, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0297, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.19555963575839996, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0321, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.1584533303976059, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0292, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.1764054298400879, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0306, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.16257385909557343, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0298, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.13896049559116364, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0296, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.16509538888931274, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0321, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.14912301301956177, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0285, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.18058481812477112, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0302, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.18734610080718994, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0289, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.19707795977592468, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0327, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.16215947270393372, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0302, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1710345447063446, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0332, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.19638675451278687, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0316, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.17807847261428833, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0285, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.13119569420814514, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0298, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.13414712250232697, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0301, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.1941734254360199, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0327, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.19624412059783936, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0291, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.17615875601768494, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0293, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.19011878967285156, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0316, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.15080299973487854, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0263, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.1676572859287262, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0318, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.14975300431251526, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0263, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17691564559936523, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.17775340378284454, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0298, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.14438258111476898, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.14729741215705872, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0289, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.20924119651317596, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0297, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.15471352636814117, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0289, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.1571032702922821, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0307, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15048687160015106, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0301, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.15063059329986572, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0279, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15528838336467743, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0299, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.1756942719221115, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0291, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.1491880565881729, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0296, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.1562451720237732, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0302, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.12366753816604614, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0257, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.15308785438537598, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.16614912450313568, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0301, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.12428635358810425, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0303, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.135500967502594, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0283, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.16944879293441772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0279, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.174363911151886, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0295, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.16406473517417908, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0284, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.2144644856452942, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0277, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.22006124258041382, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0264, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.15261462330818176, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0298, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.15205760300159454, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.14101643860340118, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0305, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16015839576721191, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.17772945761680603, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0312, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16818812489509583, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0283, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1448637843132019, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0325, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.14702047407627106, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0303, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.15547925233840942, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0299, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.18527188897132874, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.029, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.1533391922712326, + "learning_rate": 2.907944011202334e-05, + "loss": 0.029, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.15355326235294342, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0286, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.1503538340330124, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0312, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.12572433054447174, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0277, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.13614071905612946, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0303, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.13363786041736603, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0292, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.13318102061748505, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.0289, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.14687903225421906, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0301, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.13144652545452118, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0271, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.16351011395454407, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0292, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.19256997108459473, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0268, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13006120920181274, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0285, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.12932874262332916, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.11981740593910217, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.14302268624305725, + "learning_rate": 2.905937392537971e-05, + "loss": 0.029, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.1446913331747055, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0289, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.1500432789325714, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.15604284405708313, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0283, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.1795264482498169, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0289, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.14356553554534912, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0304, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.1323058009147644, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0302, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.2098175585269928, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0297, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.1424631029367447, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0295, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.1956447958946228, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0294, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.11993541568517685, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0273, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.18414588272571564, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0316, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.17320625483989716, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0319, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.17650462687015533, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0304, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.18740026652812958, + "learning_rate": 2.903909846652783e-05, + "loss": 0.034, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.12898528575897217, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0309, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.11721221357584, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0297, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.1659565567970276, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0294, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.1469241976737976, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0298, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14588896930217743, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0332, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.16243143379688263, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0305, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.15857404470443726, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0318, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.14992372691631317, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0305, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.13760966062545776, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0299, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.14302772283554077, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0313, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.18560320138931274, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0262, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.13117150962352753, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0286, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.16370882093906403, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0298, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.18689852952957153, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.031, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.16412676870822906, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0289, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.20153313875198364, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0297, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.15255704522132874, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0316, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.2036312222480774, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0306, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.16087041795253754, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0315, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.1514595001935959, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0295, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.13961558043956757, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0259, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1996241807937622, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0336, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.15389376878738403, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0278, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.12345903366804123, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.032, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.16360828280448914, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0292, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.19773486256599426, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.032, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.11931311339139938, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0268, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.1378420740365982, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0296, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.09076640754938126, + "learning_rate": 2.899643488777395e-05, + "loss": 0.026, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.1116260215640068, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0266, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.18670536577701569, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0314, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.15351636707782745, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0279, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.15310993790626526, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0298, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.15458840131759644, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0329, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.12034284323453903, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0266, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.14272604882717133, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0296, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.16335424780845642, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0266, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.16105882823467255, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0355, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.1455371230840683, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0261, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.21555180847644806, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0301, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.15278705954551697, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0292, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.14099596440792084, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0287, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.10848120599985123, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.028, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.1380046308040619, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0273, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.14006251096725464, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0328, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.16518674790859222, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0286, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.13295377790927887, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.031, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.1623251736164093, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0281, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.16029980778694153, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.03, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.15575948357582092, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0294, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.157670795917511, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0294, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1470523178577423, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0281, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.12192463129758835, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0248, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.1240374892950058, + "learning_rate": 2.895893835994484e-05, + "loss": 0.028, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1385754942893982, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0287, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.140727236866951, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0281, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.15758906304836273, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0294, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.19664499163627625, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0277, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.12567110359668732, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0279, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.13595536351203918, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0313, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.14262230694293976, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0317, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1474338322877884, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0282, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.18601401150226593, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0274, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.1392786204814911, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0269, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.12022347748279572, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0262, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.17340584099292755, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0277, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.1438349187374115, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0273, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.1408676654100418, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0277, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.14853249490261078, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0291, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.20401965081691742, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.031, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.18385861814022064, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.12695786356925964, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0267, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.14115597307682037, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0291, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.15790492296218872, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0294, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.15104074776172638, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0278, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13686735928058624, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0276, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.12029985338449478, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0305, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1351509988307953, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0293, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.10609307885169983, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0285, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15056312084197998, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0292, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.17062142491340637, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0281, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.15565407276153564, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0298, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.136429563164711, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0282, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.1649797409772873, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0267, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.19815988838672638, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.032, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15660062432289124, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.03, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.1840171217918396, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0342, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.14063017070293427, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0282, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.14935986697673798, + "learning_rate": 2.890533053632388e-05, + "loss": 0.029, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.14691074192523956, + "learning_rate": 2.89037798520224e-05, + "loss": 0.033, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.14992092549800873, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0262, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.16885656118392944, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0301, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1223178282380104, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0275, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.1200852021574974, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0277, + "step": 20000 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 20000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 64, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-40000/config.json b/checkpoint-40000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d7b77e4255263f3803a4b6f94ca6831e459b70 --- /dev/null +++ b/checkpoint-40000/config.json @@ -0,0 +1,64 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/checkpoint-40000/experiment_cfg/metadata.json b/checkpoint-40000/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..300eb070963337016e16465f50ba8351ea2eb25c --- /dev/null +++ b/checkpoint-40000/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.3139495849609375, + 0.4876587688922882, + 0.7196521759033203 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6996827125549316 + ], + "mean": [ + 2.749288365724937, + -2.0455216239909983, + 0.700532551020533 + ], + "std": [ + 1.6786295897046262, + 1.312897969434244, + 0.00133751758906258 + ], + "q01": [ + -1.4241907881148037, + -5.1716547935950885, + 0.6999670898768614 + ], + "q99": [ + 6.198111545831555, + -0.5873531334104489, + 0.7038833014242587 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.23084999587450247, + 0.6238431088581847 + ], + "std": [ + 0.0, + 0.0, + 0.6556404371644047, + 0.3572252105732042 + ], + "q01": [ + 0.0, + 0.0, + -0.9999999999999999, + 1.7482354593273349e-06 + ], + "q99": [ + 0.0, + 0.0, + 0.9999999999999999, + 0.9999999999999999 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8003877401351929, + 0.9829942584037781 + ], + "min": [ + -0.37471598386764526, + -0.8472502827644348, + -0.25070279836654663 + ], + "mean": [ + 0.28667332601863704, + -0.038315473170876704, + 0.45974658906777444 + ], + "std": [ + 0.17067058828305154, + 0.23569952037784195, + 0.21950702354181487 + ], + "q01": [ + 0.015172948963784228, + -0.44450502063220615, + 0.23314444103329338 + ], + "q99": [ + 0.5609069689572412, + 0.38454179128009547, + 0.7028102736473852 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9984455108642578, + 0.9434727430343628, + 0.9062229990959167 + ], + "min": [ + -0.9999930262565613, + -0.9988337755203247, + -0.9618149995803833, + 2.1872274658107926e-07 + ], + "mean": [ + -0.25672297401336, + 0.024425335806687216, + -0.08740977164483847, + 0.16608739644018397 + ], + "std": [ + 0.7932585208392383, + 0.3102260195580416, + 0.3772472576315148, + 0.17451959300158773 + ], + "q01": [ + -0.99379006987882, + -0.5478102050038828, + -0.6101777952971636, + 0.002274085014134748 + ], + "q99": [ + 0.8804856038896288, + 0.5910611092873037, + 0.5216058042993562, + 0.5231965974012255 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011436971835792065, + -0.05664053186774254 + ], + "mean": [ + 0.031651551516052555, + -0.03161481387920421 + ], + "std": [ + 0.013119769222217526, + 0.01306078225192402 + ], + "q01": [ + 0.0060999246446777336, + -0.04062601653964198 + ], + "q99": [ + 0.04054891621517283, + -0.0059163567113456345 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.008144896753718373, + -0.00018893135146078263, + -0.0008739062727221845, + 0.0 + ], + "std": [ + 0.11030045424364411, + 0.10148082570313594, + 0.08975698373467506, + 0.0 + ], + "q01": [ + -0.07287373067581518, + -0.09446994199569948, + -0.07702818259249572, + 0.0 + ], + "q99": [ + 0.04485485414407898, + 0.09626941605965177, + 0.07610665906122742, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9221974313425939 + ], + "std": [ + 0.38670194342612674 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.384693946883099 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.01976129923017491, + -0.020870631344490034, + -0.05993476517349181 + ], + "std": [ + 0.4347024990804053, + 0.4221662976569997, + 0.3845506843044066 + ], + "q01": [ + -0.8835792939767807, + -0.9280919251713778, + -0.783361071470956 + ], + "q99": [ + 0.7622084438449307, + 0.8625125252609077, + 0.7470974753250706 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.008089673068544335, + -0.026498254627927348, + 0.0029083551743059005 + ], + "std": [ + 0.11216734503733773, + 0.13206002613798629, + 0.12615870412692864 + ], + "q01": [ + -0.2814627972846672, + -0.4342386516115439, + -0.34489755628340574 + ], + "q99": [ + 0.31879353266939386, + 0.28411797683220563, + 0.3597718671669894 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.3824228874769045 + ], + "std": [ + 0.9240015096469786 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.5597272305813592 + ] + } + } + }, + "modalities": { + "video": { + "robot0_eye_in_hand": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_left": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_right": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": true, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + }, + "end_effector_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": true, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/checkpoint-40000/model-00001-of-00002.safetensors b/checkpoint-40000/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6caa04a8d2a2b6640dd32a1b6fe9ac36dd3158c3 --- /dev/null +++ b/checkpoint-40000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e42fd607769a5bcbb006baabeb5d52dd5598c4bd8050d4a81266399a10ef11dd +size 4999367032 diff --git a/checkpoint-40000/model-00002-of-00002.safetensors b/checkpoint-40000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c7036d73356ef78b1d084b1abc3dcac377f4c1be --- /dev/null +++ b/checkpoint-40000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1241b8b4e1fd204ddf721f6f9dea25c740ec12f0f8ee78922af719fb60ee14bd +size 2586705312 diff --git a/checkpoint-40000/model.safetensors.index.json b/checkpoint-40000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/checkpoint-40000/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-40000/optimizer.pt b/checkpoint-40000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..26f66dcbbf5d6ff73c25d58e3ac86d876d3f5584 --- /dev/null +++ b/checkpoint-40000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:305dbdd52c44d632fa786b468c57cf6aa70603bb158537b04e4c7a72fd9acb36 +size 8550719959 diff --git a/checkpoint-40000/rng_state_0.pth b/checkpoint-40000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..c959a76577d9539acf18de137c75116a475b0e9a --- /dev/null +++ b/checkpoint-40000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e6135a81c9f3879ef45b6f102d46a68b7dca4d02cb89e8c475f64fe643396655 +size 14917 diff --git a/checkpoint-40000/rng_state_1.pth b/checkpoint-40000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..3a535fcfc1f00f9723368da56b63ad7fb18043f6 --- /dev/null +++ b/checkpoint-40000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5db32635d26957c2ebda0c9fc0861ff4ab6c372497de03f9bbdbb453b63fc92f +size 14917 diff --git a/checkpoint-40000/scheduler.pt b/checkpoint-40000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..01be5087f6bfefa11717fa31027f6b4057e0ac54 --- /dev/null +++ b/checkpoint-40000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a363f689dca8e50c9cf800c4b4adeeb569ddf9651c794b79d7d723f805f5c661 +size 1465 diff --git a/checkpoint-40000/trainer_state.json b/checkpoint-40000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..100a4e8f4f43f0124cdaa8d87146c84ab7f02605 --- /dev/null +++ b/checkpoint-40000/trainer_state.json @@ -0,0 +1,28034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 11.638056444573756, + "eval_steps": 500, + "global_step": 40000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002909514111143439, + "grad_norm": 5.5908684730529785, + "learning_rate": 4.5e-08, + "loss": 0.5668, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.084368705749512, + "learning_rate": 9.5e-08, + "loss": 0.5781, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 5.755157947540283, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5759, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.174816608428955, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5202, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 6.199656963348389, + "learning_rate": 2.45e-07, + "loss": 0.5455, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.251872539520264, + "learning_rate": 2.95e-07, + "loss": 0.5507, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 5.606583595275879, + "learning_rate": 3.45e-07, + "loss": 0.5264, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.967868328094482, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.5166, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.92909574508667, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4578, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.306433200836182, + "learning_rate": 4.95e-07, + "loss": 0.4091, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.75846529006958, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4072, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.2030856609344482, + "learning_rate": 5.95e-07, + "loss": 0.2992, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.5756216049194336, + "learning_rate": 6.45e-07, + "loss": 0.2703, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.3420923948287964, + "learning_rate": 6.95e-07, + "loss": 0.2492, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 0.9191234111785889, + "learning_rate": 7.45e-07, + "loss": 0.2302, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.662022054195404, + "learning_rate": 7.95e-07, + "loss": 0.1993, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.704409658908844, + "learning_rate": 8.45e-07, + "loss": 0.1814, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.6104884147644043, + "learning_rate": 8.95e-07, + "loss": 0.1694, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.4955693781375885, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1674, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.707667887210846, + "learning_rate": 9.95e-07, + "loss": 0.1507, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.6977017521858215, + "learning_rate": 1.045e-06, + "loss": 0.1547, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.49085867404937744, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1536, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.6363996267318726, + "learning_rate": 1.145e-06, + "loss": 0.1565, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.4633055627346039, + "learning_rate": 1.195e-06, + "loss": 0.1641, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5155240297317505, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1513, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5815547108650208, + "learning_rate": 1.295e-06, + "loss": 0.1431, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.5865835547447205, + "learning_rate": 1.345e-06, + "loss": 0.1419, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.5872474312782288, + "learning_rate": 1.395e-06, + "loss": 0.1326, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.4897519052028656, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1326, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.46316635608673096, + "learning_rate": 1.495e-06, + "loss": 0.1442, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.5359047651290894, + "learning_rate": 1.545e-06, + "loss": 0.1394, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.5805708169937134, + "learning_rate": 1.595e-06, + "loss": 0.1231, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.5488415360450745, + "learning_rate": 1.645e-06, + "loss": 0.127, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.41755908727645874, + "learning_rate": 1.695e-06, + "loss": 0.1375, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.6258692145347595, + "learning_rate": 1.745e-06, + "loss": 0.1202, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.5017617344856262, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1317, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.710078239440918, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.46977895498275757, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1215, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.4802077114582062, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.116, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.6448793411254883, + "learning_rate": 1.995e-06, + "loss": 0.1172, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4394509494304657, + "learning_rate": 2.045e-06, + "loss": 0.1145, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.6547806859016418, + "learning_rate": 2.095e-06, + "loss": 0.1174, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.6954914331436157, + "learning_rate": 2.145e-06, + "loss": 0.1245, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.670100212097168, + "learning_rate": 2.195e-06, + "loss": 0.1133, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.5230806469917297, + "learning_rate": 2.245e-06, + "loss": 0.11, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.46809589862823486, + "learning_rate": 2.295e-06, + "loss": 0.1069, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.6178835034370422, + "learning_rate": 2.345e-06, + "loss": 0.1099, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.41567298769950867, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1072, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4968737065792084, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1093, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.46893924474716187, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1046, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.5041665434837341, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1014, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5833673477172852, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1095, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.5565881133079529, + "learning_rate": 2.645e-06, + "loss": 0.1057, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6213339567184448, + "learning_rate": 2.695e-06, + "loss": 0.1087, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.5638794898986816, + "learning_rate": 2.745e-06, + "loss": 0.099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.5594047904014587, + "learning_rate": 2.795e-06, + "loss": 0.1065, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5521039366722107, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1072, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.5013270378112793, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1045, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.44355231523513794, + "learning_rate": 2.945e-06, + "loss": 0.0985, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.5467342734336853, + "learning_rate": 2.995e-06, + "loss": 0.101, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5065611600875854, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0939, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.5587425827980042, + "learning_rate": 3.095e-06, + "loss": 0.0969, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5561433434486389, + "learning_rate": 3.145e-06, + "loss": 0.1009, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.5576680302619934, + "learning_rate": 3.195e-06, + "loss": 0.1006, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.7426344752311707, + "learning_rate": 3.245e-06, + "loss": 0.0959, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.64913409948349, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.0962, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.5271337628364563, + "learning_rate": 3.345e-06, + "loss": 0.0927, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.6992810368537903, + "learning_rate": 3.395e-06, + "loss": 0.0924, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.559481680393219, + "learning_rate": 3.445e-06, + "loss": 0.0886, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7303935885429382, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0918, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.45654281973838806, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0934, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.41098082065582275, + "learning_rate": 3.595e-06, + "loss": 0.0911, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.43506675958633423, + "learning_rate": 3.645e-06, + "loss": 0.0887, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5131197571754456, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0983, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.4571124017238617, + "learning_rate": 3.745e-06, + "loss": 0.0928, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.6749465465545654, + "learning_rate": 3.795e-06, + "loss": 0.0926, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.46945369243621826, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1016, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.6481541395187378, + "learning_rate": 3.895e-06, + "loss": 0.0917, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.6769195795059204, + "learning_rate": 3.945e-06, + "loss": 0.0924, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5225585103034973, + "learning_rate": 3.995e-06, + "loss": 0.0909, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.642867922782898, + "learning_rate": 4.045e-06, + "loss": 0.0829, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.4617709219455719, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0924, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.5350925326347351, + "learning_rate": 4.145e-06, + "loss": 0.0821, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.4871961176395416, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0896, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.5545480847358704, + "learning_rate": 4.245e-06, + "loss": 0.0891, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.6057930588722229, + "learning_rate": 4.295e-06, + "loss": 0.0876, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.515282154083252, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0873, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.5402254462242126, + "learning_rate": 4.395e-06, + "loss": 0.0908, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4669134318828583, + "learning_rate": 4.445e-06, + "loss": 0.0837, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.46819189190864563, + "learning_rate": 4.495e-06, + "loss": 0.0888, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35766035318374634, + "learning_rate": 4.545e-06, + "loss": 0.0829, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.6769008040428162, + "learning_rate": 4.595e-06, + "loss": 0.0817, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5884619355201721, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0892, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5613101720809937, + "learning_rate": 4.695e-06, + "loss": 0.0852, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.47087088227272034, + "learning_rate": 4.745e-06, + "loss": 0.0818, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.4887458384037018, + "learning_rate": 4.795e-06, + "loss": 0.0858, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.44256526231765747, + "learning_rate": 4.845e-06, + "loss": 0.0833, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5375688076019287, + "learning_rate": 4.895e-06, + "loss": 0.075, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.6297019124031067, + "learning_rate": 4.945e-06, + "loss": 0.0932, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.6576125621795654, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0829, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.724773108959198, + "learning_rate": 5.045e-06, + "loss": 0.0813, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.54575115442276, + "learning_rate": 5.095e-06, + "loss": 0.0873, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.5186223387718201, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0736, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.46651381254196167, + "learning_rate": 5.195e-06, + "loss": 0.0791, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.5323564410209656, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0753, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.42579638957977295, + "learning_rate": 5.295e-06, + "loss": 0.0755, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.5063076019287109, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0767, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.4558819532394409, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0807, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.6920697093009949, + "learning_rate": 5.445e-06, + "loss": 0.079, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.4706108868122101, + "learning_rate": 5.495e-06, + "loss": 0.079, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.46705493330955505, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0745, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4643767774105072, + "learning_rate": 5.595e-06, + "loss": 0.0778, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.4800759553909302, + "learning_rate": 5.645e-06, + "loss": 0.0792, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.37705323100090027, + "learning_rate": 5.695e-06, + "loss": 0.0815, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.6651142239570618, + "learning_rate": 5.745e-06, + "loss": 0.0732, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.5275036692619324, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0758, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.45047008991241455, + "learning_rate": 5.845e-06, + "loss": 0.0808, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.6931971311569214, + "learning_rate": 5.895e-06, + "loss": 0.0808, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.7053916454315186, + "learning_rate": 5.945e-06, + "loss": 0.0747, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.6798260807991028, + "learning_rate": 5.995e-06, + "loss": 0.0829, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.5724003314971924, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0749, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.5836488008499146, + "learning_rate": 6.095e-06, + "loss": 0.0759, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.40817850828170776, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0778, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.41070470213890076, + "learning_rate": 6.195e-06, + "loss": 0.0765, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.43642961978912354, + "learning_rate": 6.245e-06, + "loss": 0.0711, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.49526578187942505, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0713, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.7773705720901489, + "learning_rate": 6.345e-06, + "loss": 0.0744, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.5485936999320984, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0732, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.5593521595001221, + "learning_rate": 6.445e-06, + "loss": 0.0747, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.3667270839214325, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0723, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.5555917620658875, + "learning_rate": 6.545e-06, + "loss": 0.0726, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.38107141852378845, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0647, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.46452224254608154, + "learning_rate": 6.645e-06, + "loss": 0.0787, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6018183827400208, + "learning_rate": 6.695e-06, + "loss": 0.0718, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.44997578859329224, + "learning_rate": 6.745e-06, + "loss": 0.0758, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45717573165893555, + "learning_rate": 6.795e-06, + "loss": 0.0738, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.5185182690620422, + "learning_rate": 6.845e-06, + "loss": 0.0768, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.5135819315910339, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.43529245257377625, + "learning_rate": 6.945e-06, + "loss": 0.0691, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.49389752745628357, + "learning_rate": 6.995e-06, + "loss": 0.0678, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.4113156199455261, + "learning_rate": 7.045e-06, + "loss": 0.0651, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.4072611927986145, + "learning_rate": 7.095e-06, + "loss": 0.0691, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.4524228870868683, + "learning_rate": 7.145e-06, + "loss": 0.0688, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.6225605607032776, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0676, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4841763377189636, + "learning_rate": 7.245e-06, + "loss": 0.0678, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.46297937631607056, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0759, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.5585324168205261, + "learning_rate": 7.345e-06, + "loss": 0.0671, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.57328200340271, + "learning_rate": 7.395e-06, + "loss": 0.0717, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.5072467923164368, + "learning_rate": 7.445000000000001e-06, + "loss": 0.071, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.4630240797996521, + "learning_rate": 7.495e-06, + "loss": 0.0674, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.35908272862434387, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0704, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.5117342472076416, + "learning_rate": 7.595e-06, + "loss": 0.0675, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.5214630961418152, + "learning_rate": 7.645e-06, + "loss": 0.0702, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.39036089181900024, + "learning_rate": 7.695e-06, + "loss": 0.0709, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.558722734451294, + "learning_rate": 7.745e-06, + "loss": 0.0705, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.3612176179885864, + "learning_rate": 7.795e-06, + "loss": 0.0715, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.523444652557373, + "learning_rate": 7.845e-06, + "loss": 0.0685, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.5231291651725769, + "learning_rate": 7.895e-06, + "loss": 0.0717, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.50368332862854, + "learning_rate": 7.945e-06, + "loss": 0.0694, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.4126806855201721, + "learning_rate": 7.995e-06, + "loss": 0.0693, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.719352662563324, + "learning_rate": 8.045e-06, + "loss": 0.0651, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.4405677020549774, + "learning_rate": 8.095e-06, + "loss": 0.0645, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.34746286273002625, + "learning_rate": 8.145e-06, + "loss": 0.0687, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.43200597167015076, + "learning_rate": 8.195e-06, + "loss": 0.0655, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.3464098274707794, + "learning_rate": 8.245e-06, + "loss": 0.0639, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.27263233065605164, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0629, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.343181848526001, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3949901759624481, + "learning_rate": 8.395e-06, + "loss": 0.0643, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.4489574730396271, + "learning_rate": 8.445e-06, + "loss": 0.0698, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.4746992588043213, + "learning_rate": 8.495e-06, + "loss": 0.0702, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.3651655316352844, + "learning_rate": 8.545e-06, + "loss": 0.0673, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.471071720123291, + "learning_rate": 8.595e-06, + "loss": 0.0648, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.35623008012771606, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0698, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.41837432980537415, + "learning_rate": 8.695e-06, + "loss": 0.0639, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.45608940720558167, + "learning_rate": 8.745e-06, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.40346869826316833, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0625, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.479824960231781, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0703, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.38570791482925415, + "learning_rate": 8.895e-06, + "loss": 0.0643, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.36530473828315735, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0652, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37868866324424744, + "learning_rate": 8.995000000000001e-06, + "loss": 0.068, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.32546791434288025, + "learning_rate": 9.045e-06, + "loss": 0.0636, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4664970338344574, + "learning_rate": 9.095e-06, + "loss": 0.0671, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4767569601535797, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0625, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.41823142766952515, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0685, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.3709985613822937, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0681, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.3933296799659729, + "learning_rate": 9.295e-06, + "loss": 0.0678, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.3079732060432434, + "learning_rate": 9.345e-06, + "loss": 0.0598, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.33485978841781616, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0644, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.40312913060188293, + "learning_rate": 9.445e-06, + "loss": 0.0684, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.4334392547607422, + "learning_rate": 9.495e-06, + "loss": 0.0622, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.5787541270256042, + "learning_rate": 9.545e-06, + "loss": 0.0656, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.2827294170856476, + "learning_rate": 9.595e-06, + "loss": 0.0628, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4431678056716919, + "learning_rate": 9.645e-06, + "loss": 0.0633, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.3916102647781372, + "learning_rate": 9.695e-06, + "loss": 0.0632, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.41470348834991455, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0612, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.3224067986011505, + "learning_rate": 9.795e-06, + "loss": 0.061, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.4110186696052551, + "learning_rate": 9.845e-06, + "loss": 0.066, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.4786360561847687, + "learning_rate": 9.895e-06, + "loss": 0.0648, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3667179048061371, + "learning_rate": 9.945e-06, + "loss": 0.0663, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.31722185015678406, + "learning_rate": 9.995e-06, + "loss": 0.0588, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.29271745681762695, + "learning_rate": 1.0045e-05, + "loss": 0.0636, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.3536592423915863, + "learning_rate": 1.0095e-05, + "loss": 0.0625, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.3228510916233063, + "learning_rate": 1.0145e-05, + "loss": 0.067, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.38474392890930176, + "learning_rate": 1.0195e-05, + "loss": 0.0609, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.40410760045051575, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0638, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.3632320463657379, + "learning_rate": 1.0295e-05, + "loss": 0.0587, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.37775692343711853, + "learning_rate": 1.0345e-05, + "loss": 0.0673, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.3343006372451782, + "learning_rate": 1.0395e-05, + "loss": 0.0599, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3232511281967163, + "learning_rate": 1.0445e-05, + "loss": 0.0661, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.38730430603027344, + "learning_rate": 1.0495e-05, + "loss": 0.0649, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.38972100615501404, + "learning_rate": 1.0545e-05, + "loss": 0.0582, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.27242428064346313, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0583, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.30514073371887207, + "learning_rate": 1.0645e-05, + "loss": 0.065, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.45364341139793396, + "learning_rate": 1.0695e-05, + "loss": 0.0576, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.3863564729690552, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0579, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.33717092871665955, + "learning_rate": 1.0795e-05, + "loss": 0.0565, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38073790073394775, + "learning_rate": 1.0845e-05, + "loss": 0.0585, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4277617037296295, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0622, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.3545698821544647, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0609, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.30646640062332153, + "learning_rate": 1.0995e-05, + "loss": 0.0659, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.3607352375984192, + "learning_rate": 1.1045e-05, + "loss": 0.0621, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3624425530433655, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0626, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3856692612171173, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0604, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.4390200972557068, + "learning_rate": 1.1195e-05, + "loss": 0.0601, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.4108872711658478, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0618, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.32732832431793213, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0636, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.3602239787578583, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.29978248476982117, + "learning_rate": 1.1395e-05, + "loss": 0.058, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.48490920662879944, + "learning_rate": 1.1445e-05, + "loss": 0.0614, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.3278113901615143, + "learning_rate": 1.1495e-05, + "loss": 0.0611, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.37862712144851685, + "learning_rate": 1.1545e-05, + "loss": 0.0626, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.3514714241027832, + "learning_rate": 1.1595e-05, + "loss": 0.0583, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.27256545424461365, + "learning_rate": 1.1645e-05, + "loss": 0.062, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.3305833339691162, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0572, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.2881031930446625, + "learning_rate": 1.1745e-05, + "loss": 0.0641, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.3219377398490906, + "learning_rate": 1.1795e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.34189504384994507, + "learning_rate": 1.1845e-05, + "loss": 0.0607, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30541297793388367, + "learning_rate": 1.1895e-05, + "loss": 0.056, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.35898879170417786, + "learning_rate": 1.1945e-05, + "loss": 0.0598, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.33221277594566345, + "learning_rate": 1.1995e-05, + "loss": 0.0603, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.3435590863227844, + "learning_rate": 1.2045e-05, + "loss": 0.0553, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.24820426106452942, + "learning_rate": 1.2095e-05, + "loss": 0.0644, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.3749466836452484, + "learning_rate": 1.2145e-05, + "loss": 0.0615, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.3984697759151459, + "learning_rate": 1.2195e-05, + "loss": 0.0595, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3250048756599426, + "learning_rate": 1.2245e-05, + "loss": 0.0608, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.2794302701950073, + "learning_rate": 1.2295e-05, + "loss": 0.0578, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3811834752559662, + "learning_rate": 1.2345e-05, + "loss": 0.0612, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3160156309604645, + "learning_rate": 1.2395e-05, + "loss": 0.0541, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.3825189173221588, + "learning_rate": 1.2445e-05, + "loss": 0.0621, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.3966415226459503, + "learning_rate": 1.2495e-05, + "loss": 0.0598, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3615199625492096, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0616, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.440762996673584, + "learning_rate": 1.2595e-05, + "loss": 0.0599, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.345599889755249, + "learning_rate": 1.2645e-05, + "loss": 0.0597, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.2963278889656067, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0556, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.3720734417438507, + "learning_rate": 1.2745e-05, + "loss": 0.0618, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3937445878982544, + "learning_rate": 1.2795e-05, + "loss": 0.0624, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.4492146372795105, + "learning_rate": 1.2845e-05, + "loss": 0.0573, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.5172158479690552, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0594, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.3796320855617523, + "learning_rate": 1.2945e-05, + "loss": 0.0528, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.42423033714294434, + "learning_rate": 1.2995e-05, + "loss": 0.0615, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.4578545093536377, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0562, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.38345491886138916, + "learning_rate": 1.3095e-05, + "loss": 0.0574, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.2539816200733185, + "learning_rate": 1.3145e-05, + "loss": 0.0562, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.44323891401290894, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0588, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.4074377417564392, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0591, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.5252887606620789, + "learning_rate": 1.3295e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.39536044001579285, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0588, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.6300645470619202, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0584, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.312466561794281, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0588, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3810383379459381, + "learning_rate": 1.3495e-05, + "loss": 0.0597, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.30645400285720825, + "learning_rate": 1.3545e-05, + "loss": 0.0589, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.2857605218887329, + "learning_rate": 1.3595e-05, + "loss": 0.0589, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.3206622898578644, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0597, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3226073086261749, + "learning_rate": 1.3695e-05, + "loss": 0.0559, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.24912619590759277, + "learning_rate": 1.3745e-05, + "loss": 0.0547, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.29700762033462524, + "learning_rate": 1.3795e-05, + "loss": 0.0563, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.32627588510513306, + "learning_rate": 1.3845e-05, + "loss": 0.0602, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.48569780588150024, + "learning_rate": 1.3895e-05, + "loss": 0.0553, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.323607474565506, + "learning_rate": 1.3945e-05, + "loss": 0.0578, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.42626142501831055, + "learning_rate": 1.3995e-05, + "loss": 0.0539, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.3281888961791992, + "learning_rate": 1.4045e-05, + "loss": 0.0638, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.38216397166252136, + "learning_rate": 1.4095e-05, + "loss": 0.0572, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.2816051244735718, + "learning_rate": 1.4145e-05, + "loss": 0.0543, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.2604160010814667, + "learning_rate": 1.4195e-05, + "loss": 0.0555, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.34318771958351135, + "learning_rate": 1.4245e-05, + "loss": 0.0565, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.31248190999031067, + "learning_rate": 1.4295e-05, + "loss": 0.0583, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.3014426827430725, + "learning_rate": 1.4345e-05, + "loss": 0.0586, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3650360703468323, + "learning_rate": 1.4395e-05, + "loss": 0.0581, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.33519116044044495, + "learning_rate": 1.4445e-05, + "loss": 0.0611, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.3595008850097656, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.054, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.32799217104911804, + "learning_rate": 1.4545e-05, + "loss": 0.0574, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.29410457611083984, + "learning_rate": 1.4595e-05, + "loss": 0.0582, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.2727576792240143, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0588, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.36562401056289673, + "learning_rate": 1.4695e-05, + "loss": 0.0597, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.3351168930530548, + "learning_rate": 1.4745e-05, + "loss": 0.0558, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.23880337178707123, + "learning_rate": 1.4795e-05, + "loss": 0.0571, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.3910059928894043, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0573, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.3103587031364441, + "learning_rate": 1.4895e-05, + "loss": 0.0578, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.3265620768070221, + "learning_rate": 1.4945e-05, + "loss": 0.0542, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.3060731291770935, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0537, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.41589322686195374, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0546, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.3102424740791321, + "learning_rate": 1.5095e-05, + "loss": 0.0564, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.32286810874938965, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0519, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.3129369020462036, + "learning_rate": 1.5195e-05, + "loss": 0.0566, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.43735018372535706, + "learning_rate": 1.5245e-05, + "loss": 0.0552, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.440449595451355, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0589, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.3631616532802582, + "learning_rate": 1.5345e-05, + "loss": 0.0511, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.29124221205711365, + "learning_rate": 1.5395e-05, + "loss": 0.055, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.4247622489929199, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0562, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.36976534128189087, + "learning_rate": 1.5495e-05, + "loss": 0.0604, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.3580514192581177, + "learning_rate": 1.5545e-05, + "loss": 0.0547, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.2969493567943573, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0564, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28583595156669617, + "learning_rate": 1.5645e-05, + "loss": 0.0586, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.3220328390598297, + "learning_rate": 1.5695e-05, + "loss": 0.0588, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.32270652055740356, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0559, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.3170712888240814, + "learning_rate": 1.5795e-05, + "loss": 0.0566, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.2966325879096985, + "learning_rate": 1.5845e-05, + "loss": 0.0551, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.2788766622543335, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.051, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.3521708548069, + "learning_rate": 1.5945e-05, + "loss": 0.0558, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.35105040669441223, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0551, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.4221852123737335, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0573, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.4217511713504791, + "learning_rate": 1.6095e-05, + "loss": 0.0532, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.4519175589084625, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0489, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3427059054374695, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0544, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29204127192497253, + "learning_rate": 1.6245e-05, + "loss": 0.0525, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.24138635396957397, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0559, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.3599991500377655, + "learning_rate": 1.6345e-05, + "loss": 0.0544, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.3326258361339569, + "learning_rate": 1.6395e-05, + "loss": 0.0593, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31976261734962463, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0624, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.26379555463790894, + "learning_rate": 1.6495e-05, + "loss": 0.0529, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2576923966407776, + "learning_rate": 1.6545e-05, + "loss": 0.0527, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.30890268087387085, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0571, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.4547998309135437, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.3050350248813629, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0552, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.2984176278114319, + "learning_rate": 1.6745e-05, + "loss": 0.0528, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.34583792090415955, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0512, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.33085328340530396, + "learning_rate": 1.6845e-05, + "loss": 0.0563, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.33391326665878296, + "learning_rate": 1.6895e-05, + "loss": 0.0527, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.23657914996147156, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0509, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.3719659149646759, + "learning_rate": 1.6995e-05, + "loss": 0.0568, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.4268389940261841, + "learning_rate": 1.7045e-05, + "loss": 0.0525, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.2726958394050598, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0482, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27847009897232056, + "learning_rate": 1.7145e-05, + "loss": 0.0559, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.7195e-05, + "loss": 0.0588, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.2666028141975403, + "learning_rate": 1.7245e-05, + "loss": 0.0548, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.26048436760902405, + "learning_rate": 1.7295e-05, + "loss": 0.0592, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.22569318115711212, + "learning_rate": 1.7345e-05, + "loss": 0.0505, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.35350096225738525, + "learning_rate": 1.7395e-05, + "loss": 0.0566, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.3166314363479614, + "learning_rate": 1.7445e-05, + "loss": 0.0517, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.24918541312217712, + "learning_rate": 1.7495e-05, + "loss": 0.0545, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3416207730770111, + "learning_rate": 1.7545e-05, + "loss": 0.0544, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.22633257508277893, + "learning_rate": 1.7595e-05, + "loss": 0.0545, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.3972736597061157, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0591, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3213656544685364, + "learning_rate": 1.7695e-05, + "loss": 0.0537, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.30361926555633545, + "learning_rate": 1.7745e-05, + "loss": 0.0529, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.35521000623703003, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0549, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.4444983899593353, + "learning_rate": 1.7845e-05, + "loss": 0.0529, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.32495826482772827, + "learning_rate": 1.7895e-05, + "loss": 0.0557, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.2893896698951721, + "learning_rate": 1.7945e-05, + "loss": 0.056, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.2607886493206024, + "learning_rate": 1.7995e-05, + "loss": 0.055, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.3305937647819519, + "learning_rate": 1.8045e-05, + "loss": 0.0529, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.3003319799900055, + "learning_rate": 1.8095e-05, + "loss": 0.0515, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3523712456226349, + "learning_rate": 1.8145e-05, + "loss": 0.0524, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.27575406432151794, + "learning_rate": 1.8195e-05, + "loss": 0.0535, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.36445286870002747, + "learning_rate": 1.8245e-05, + "loss": 0.0552, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.41415491700172424, + "learning_rate": 1.8295e-05, + "loss": 0.0537, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.3738953471183777, + "learning_rate": 1.8345e-05, + "loss": 0.0555, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.3446105420589447, + "learning_rate": 1.8395e-05, + "loss": 0.0526, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.39016082882881165, + "learning_rate": 1.8445e-05, + "loss": 0.0536, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.4075721502304077, + "learning_rate": 1.8495e-05, + "loss": 0.0548, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.4071345925331116, + "learning_rate": 1.8545e-05, + "loss": 0.0539, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.34446755051612854, + "learning_rate": 1.8595e-05, + "loss": 0.057, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.3431609272956848, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0515, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.31537994742393494, + "learning_rate": 1.8695e-05, + "loss": 0.0534, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.4846970736980438, + "learning_rate": 1.8745e-05, + "loss": 0.0532, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.3109683394432068, + "learning_rate": 1.8795e-05, + "loss": 0.0547, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.28903284668922424, + "learning_rate": 1.8845e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.26592350006103516, + "learning_rate": 1.8895e-05, + "loss": 0.0499, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.27347302436828613, + "learning_rate": 1.8945e-05, + "loss": 0.052, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.3194405138492584, + "learning_rate": 1.8995e-05, + "loss": 0.0521, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.40457430481910706, + "learning_rate": 1.9045e-05, + "loss": 0.0523, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.3055919408798218, + "learning_rate": 1.9095e-05, + "loss": 0.0504, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.4911873936653137, + "learning_rate": 1.9145e-05, + "loss": 0.0531, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.30619072914123535, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0525, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.24404728412628174, + "learning_rate": 1.9245e-05, + "loss": 0.0498, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.31723177433013916, + "learning_rate": 1.9295e-05, + "loss": 0.0519, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.28105342388153076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0528, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.32965341210365295, + "learning_rate": 1.9395e-05, + "loss": 0.0504, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.2487417608499527, + "learning_rate": 1.9445e-05, + "loss": 0.0552, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3788810670375824, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0535, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.2561226189136505, + "learning_rate": 1.9545e-05, + "loss": 0.056, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.29964715242385864, + "learning_rate": 1.9595e-05, + "loss": 0.0565, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.386600136756897, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0468, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.37690475583076477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.35367539525032043, + "learning_rate": 1.9745e-05, + "loss": 0.0501, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.30189844965934753, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0524, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.27525416016578674, + "learning_rate": 1.9845e-05, + "loss": 0.0524, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.22557462751865387, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.3572654128074646, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0536, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.37130483984947205, + "learning_rate": 1.9995e-05, + "loss": 0.0507, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.40820422768592834, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0547, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.24222403764724731, + "learning_rate": 2.0095e-05, + "loss": 0.0536, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.35473307967185974, + "learning_rate": 2.0145e-05, + "loss": 0.0503, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24238434433937073, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0511, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.32721346616744995, + "learning_rate": 2.0245e-05, + "loss": 0.0474, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.3787407875061035, + "learning_rate": 2.0295e-05, + "loss": 0.0515, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.3435424566268921, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0521, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.250515341758728, + "learning_rate": 2.0395e-05, + "loss": 0.0539, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.2730806767940521, + "learning_rate": 2.0445e-05, + "loss": 0.0544, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.32704228162765503, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0516, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.26865822076797485, + "learning_rate": 2.0545e-05, + "loss": 0.0557, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.31514087319374084, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.051, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.2616795301437378, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0483, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2979176938533783, + "learning_rate": 2.0695e-05, + "loss": 0.056, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.2876877188682556, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0496, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.2801620662212372, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0533, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.3521801233291626, + "learning_rate": 2.0845e-05, + "loss": 0.051, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.3016152083873749, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0515, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.29837727546691895, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0483, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.255982369184494, + "learning_rate": 2.0995e-05, + "loss": 0.0482, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.20852255821228027, + "learning_rate": 2.1045e-05, + "loss": 0.05, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.2746782898902893, + "learning_rate": 2.1095e-05, + "loss": 0.0513, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.28745001554489136, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0486, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.3163246810436249, + "learning_rate": 2.1195e-05, + "loss": 0.0536, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.2279432713985443, + "learning_rate": 2.1245e-05, + "loss": 0.0497, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.2273164689540863, + "learning_rate": 2.1295e-05, + "loss": 0.0485, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.3124007284641266, + "learning_rate": 2.1345e-05, + "loss": 0.0473, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.2097717970609665, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0514, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.24428582191467285, + "learning_rate": 2.1445e-05, + "loss": 0.0543, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.47236987948417664, + "learning_rate": 2.1495e-05, + "loss": 0.0525, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.3367164433002472, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0525, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.34086376428604126, + "learning_rate": 2.1595e-05, + "loss": 0.049, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.3060646951198578, + "learning_rate": 2.1645e-05, + "loss": 0.0538, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.2997030019760132, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0553, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.2652549743652344, + "learning_rate": 2.1745e-05, + "loss": 0.051, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.3072361350059509, + "learning_rate": 2.1795e-05, + "loss": 0.0516, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2494215965270996, + "learning_rate": 2.1845e-05, + "loss": 0.0476, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.20796196162700653, + "learning_rate": 2.1895e-05, + "loss": 0.0553, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.24235095083713531, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.21033242344856262, + "learning_rate": 2.1995e-05, + "loss": 0.0542, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.26056164503097534, + "learning_rate": 2.2045e-05, + "loss": 0.0546, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.24798662960529327, + "learning_rate": 2.2095e-05, + "loss": 0.0529, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.22179707884788513, + "learning_rate": 2.2145e-05, + "loss": 0.0542, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.2660025358200073, + "learning_rate": 2.2195e-05, + "loss": 0.0496, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.3771500289440155, + "learning_rate": 2.2245e-05, + "loss": 0.0502, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.3059975206851959, + "learning_rate": 2.2295e-05, + "loss": 0.0482, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.2941029667854309, + "learning_rate": 2.2345e-05, + "loss": 0.0483, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3481833040714264, + "learning_rate": 2.2395e-05, + "loss": 0.0541, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.2729704678058624, + "learning_rate": 2.2445e-05, + "loss": 0.0514, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.44202038645744324, + "learning_rate": 2.2495e-05, + "loss": 0.0534, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.311088889837265, + "learning_rate": 2.2545e-05, + "loss": 0.0474, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30663055181503296, + "learning_rate": 2.2595e-05, + "loss": 0.0491, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.3530942499637604, + "learning_rate": 2.2645e-05, + "loss": 0.0503, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.27998727560043335, + "learning_rate": 2.2695e-05, + "loss": 0.0493, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.2679221034049988, + "learning_rate": 2.2745e-05, + "loss": 0.0512, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.348821222782135, + "learning_rate": 2.2795e-05, + "loss": 0.0489, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.260958194732666, + "learning_rate": 2.2845e-05, + "loss": 0.0474, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.2429468184709549, + "learning_rate": 2.2895e-05, + "loss": 0.053, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.2635575234889984, + "learning_rate": 2.2945e-05, + "loss": 0.053, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.3416514992713928, + "learning_rate": 2.2995e-05, + "loss": 0.0476, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.221172496676445, + "learning_rate": 2.3045e-05, + "loss": 0.0519, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.30541878938674927, + "learning_rate": 2.3095e-05, + "loss": 0.0522, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.25274574756622314, + "learning_rate": 2.3145e-05, + "loss": 0.0499, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.366253525018692, + "learning_rate": 2.3195e-05, + "loss": 0.0463, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.16825024783611298, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0493, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.3158467710018158, + "learning_rate": 2.3295e-05, + "loss": 0.0482, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.2887287139892578, + "learning_rate": 2.3345e-05, + "loss": 0.053, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.4717058539390564, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0515, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.3148113489151001, + "learning_rate": 2.3445e-05, + "loss": 0.0523, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.23815515637397766, + "learning_rate": 2.3495e-05, + "loss": 0.0523, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23897255957126617, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0484, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.2680213749408722, + "learning_rate": 2.3595e-05, + "loss": 0.0485, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.29999756813049316, + "learning_rate": 2.3645e-05, + "loss": 0.0506, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.30607151985168457, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.05, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.26077136397361755, + "learning_rate": 2.3745e-05, + "loss": 0.0558, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2638765275478363, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.049, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.24569877982139587, + "learning_rate": 2.3845e-05, + "loss": 0.053, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.29622989892959595, + "learning_rate": 2.3895e-05, + "loss": 0.0517, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.2598831355571747, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.05, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.25716227293014526, + "learning_rate": 2.3995e-05, + "loss": 0.0453, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.37305381894111633, + "learning_rate": 2.4045e-05, + "loss": 0.0539, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.3151630461215973, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0516, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.296941339969635, + "learning_rate": 2.4145e-05, + "loss": 0.0509, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.31995704770088196, + "learning_rate": 2.4195e-05, + "loss": 0.0498, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2426208257675171, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0475, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.23927678167819977, + "learning_rate": 2.4295e-05, + "loss": 0.05, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.1887904703617096, + "learning_rate": 2.4345e-05, + "loss": 0.05, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.18480589985847473, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0496, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.24515867233276367, + "learning_rate": 2.4445e-05, + "loss": 0.0464, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.25505101680755615, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0481, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.24435125291347504, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0502, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.25899314880371094, + "learning_rate": 2.4595e-05, + "loss": 0.0535, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.30518922209739685, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0481, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.26098036766052246, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0477, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.24723340570926666, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.26713746786117554, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0506, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.21166373789310455, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0502, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2602159082889557, + "learning_rate": 2.4895e-05, + "loss": 0.0506, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27918240427970886, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.049, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.2131885290145874, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0477, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.28778040409088135, + "learning_rate": 2.5045e-05, + "loss": 0.0487, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.3056010603904724, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.049, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.2850024700164795, + "learning_rate": 2.5145e-05, + "loss": 0.0475, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.20490600168704987, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0497, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.32371339201927185, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0491, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.32466208934783936, + "learning_rate": 2.5295e-05, + "loss": 0.0485, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.2779853343963623, + "learning_rate": 2.5345e-05, + "loss": 0.0485, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.3257012367248535, + "learning_rate": 2.5395e-05, + "loss": 0.0486, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2625649571418762, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0486, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.19846542179584503, + "learning_rate": 2.5495e-05, + "loss": 0.0494, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.2415613830089569, + "learning_rate": 2.5545e-05, + "loss": 0.0539, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.20438073575496674, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0476, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.2543872892856598, + "learning_rate": 2.5645e-05, + "loss": 0.0489, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.26969635486602783, + "learning_rate": 2.5695e-05, + "loss": 0.0557, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.27818888425827026, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0517, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.28826993703842163, + "learning_rate": 2.5795e-05, + "loss": 0.0499, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.19518250226974487, + "learning_rate": 2.5845e-05, + "loss": 0.0499, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.2639068067073822, + "learning_rate": 2.5895e-05, + "loss": 0.0459, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.26014742255210876, + "learning_rate": 2.5945e-05, + "loss": 0.0522, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.26659756898880005, + "learning_rate": 2.5995e-05, + "loss": 0.0471, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.24073593318462372, + "learning_rate": 2.6045e-05, + "loss": 0.0488, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.2724616527557373, + "learning_rate": 2.6095e-05, + "loss": 0.0493, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.3012109398841858, + "learning_rate": 2.6145e-05, + "loss": 0.049, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23306235671043396, + "learning_rate": 2.6195e-05, + "loss": 0.0474, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.23879121243953705, + "learning_rate": 2.6245e-05, + "loss": 0.048, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20701922476291656, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0527, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.28734415769577026, + "learning_rate": 2.6345e-05, + "loss": 0.0469, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.26227426528930664, + "learning_rate": 2.6395e-05, + "loss": 0.0472, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.24196277558803558, + "learning_rate": 2.6445e-05, + "loss": 0.0462, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.3405964970588684, + "learning_rate": 2.6495e-05, + "loss": 0.0479, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.24121852219104767, + "learning_rate": 2.6545e-05, + "loss": 0.0444, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.284856379032135, + "learning_rate": 2.6595e-05, + "loss": 0.0492, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.2515311539173126, + "learning_rate": 2.6645e-05, + "loss": 0.0453, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.29297393560409546, + "learning_rate": 2.6695e-05, + "loss": 0.049, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.2528100311756134, + "learning_rate": 2.6745e-05, + "loss": 0.0477, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.29126009345054626, + "learning_rate": 2.6795e-05, + "loss": 0.0484, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.26489511132240295, + "learning_rate": 2.6845e-05, + "loss": 0.0476, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.33150023221969604, + "learning_rate": 2.6895e-05, + "loss": 0.0475, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.25904157757759094, + "learning_rate": 2.6945e-05, + "loss": 0.0495, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.24830900132656097, + "learning_rate": 2.6995e-05, + "loss": 0.0476, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.277997225522995, + "learning_rate": 2.7045e-05, + "loss": 0.0478, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.3100328743457794, + "learning_rate": 2.7095e-05, + "loss": 0.0501, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.2957077622413635, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0491, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.19402435421943665, + "learning_rate": 2.7195e-05, + "loss": 0.0426, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.20080482959747314, + "learning_rate": 2.7245e-05, + "loss": 0.0464, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.21148671209812164, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0457, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.2299162596464157, + "learning_rate": 2.7345e-05, + "loss": 0.0467, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2296467125415802, + "learning_rate": 2.7395e-05, + "loss": 0.0489, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.25099772214889526, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0487, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2503034472465515, + "learning_rate": 2.7495e-05, + "loss": 0.0533, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.24678738415241241, + "learning_rate": 2.7545e-05, + "loss": 0.0462, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.3281477987766266, + "learning_rate": 2.7595e-05, + "loss": 0.0492, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.32214412093162537, + "learning_rate": 2.7645e-05, + "loss": 0.0459, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.29244086146354675, + "learning_rate": 2.7695e-05, + "loss": 0.0495, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.2372455894947052, + "learning_rate": 2.7745e-05, + "loss": 0.0477, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.28411489725112915, + "learning_rate": 2.7795e-05, + "loss": 0.0514, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.36217001080513, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2913915514945984, + "learning_rate": 2.7895e-05, + "loss": 0.0535, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.3237503170967102, + "learning_rate": 2.7945e-05, + "loss": 0.0509, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.32016751170158386, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0526, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.25484171509742737, + "learning_rate": 2.8045e-05, + "loss": 0.0508, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.22641469538211823, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.23269280791282654, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0484, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23346202075481415, + "learning_rate": 2.8195e-05, + "loss": 0.0472, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.25522616505622864, + "learning_rate": 2.8245e-05, + "loss": 0.0493, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.2478165179491043, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0468, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.25732970237731934, + "learning_rate": 2.8345e-05, + "loss": 0.0484, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.21047812700271606, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0455, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.252676397562027, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0496, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20048962533473969, + "learning_rate": 2.8495e-05, + "loss": 0.0434, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.1821785420179367, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0502, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.24106337130069733, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0471, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.413974404335022, + "learning_rate": 2.8645e-05, + "loss": 0.0542, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.29419073462486267, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0499, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.20694036781787872, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0453, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2390667200088501, + "learning_rate": 2.8795e-05, + "loss": 0.045, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2563699781894684, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.046, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.22678422927856445, + "learning_rate": 2.8895e-05, + "loss": 0.0456, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.26104822754859924, + "learning_rate": 2.8945e-05, + "loss": 0.0446, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.22175569832324982, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0478, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.20990405976772308, + "learning_rate": 2.9045e-05, + "loss": 0.0459, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.27277135848999023, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0477, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.22422750294208527, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.047, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.2740926742553711, + "learning_rate": 2.9195e-05, + "loss": 0.0444, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2977927327156067, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0477, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26431694626808167, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0488, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.22759316861629486, + "learning_rate": 2.9345e-05, + "loss": 0.0465, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.24671563506126404, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0501, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.24065586924552917, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0439, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.3334764838218689, + "learning_rate": 2.9495e-05, + "loss": 0.0475, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.23565605282783508, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.046, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.26102808117866516, + "learning_rate": 2.9595e-05, + "loss": 0.0494, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.34544840455055237, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.05, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.2476256936788559, + "learning_rate": 2.9695e-05, + "loss": 0.0477, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.2795628309249878, + "learning_rate": 2.9745e-05, + "loss": 0.0508, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.2756715714931488, + "learning_rate": 2.9795e-05, + "loss": 0.0484, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.20578329265117645, + "learning_rate": 2.9845e-05, + "loss": 0.0461, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.20354008674621582, + "learning_rate": 2.9895e-05, + "loss": 0.0431, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24710936844348907, + "learning_rate": 2.9945e-05, + "loss": 0.0459, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.32639119029045105, + "learning_rate": 2.9995e-05, + "loss": 0.045, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.27762553095817566, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0476, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.2903423607349396, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.3292105793952942, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0503, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.23503656685352325, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0446, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.24153585731983185, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.048, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.24157844483852386, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.046, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.23378467559814453, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0478, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.20753100514411926, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0527, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.2574090361595154, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0468, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.23341813683509827, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0445, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21378666162490845, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0487, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.2410328984260559, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0488, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.22005778551101685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0478, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.3197212219238281, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.046, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.253745436668396, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0437, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.1991635262966156, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0474, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.20773734152317047, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0439, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.3026197552680969, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0462, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.22027666866779327, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0443, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.2512618899345398, + "learning_rate": 2.99997744429684e-05, + "loss": 0.044, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.23001381754875183, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0479, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.20057332515716553, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0433, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.2831740975379944, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0462, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.18362601101398468, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0472, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.29741403460502625, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0443, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.20039376616477966, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0457, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.14999084174633026, + "learning_rate": 2.999958785135414e-05, + "loss": 0.043, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.18134453892707825, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0416, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.25010257959365845, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0494, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.31700536608695984, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0455, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.25080761313438416, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0436, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.2008715718984604, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0459, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.21441133320331573, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0492, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.23232029378414154, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0459, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.30173608660697937, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0505, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.21179363131523132, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0438, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.21905283629894257, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0439, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.17932383716106415, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0452, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.21132917702198029, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0454, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.22732721269130707, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.047, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.28214913606643677, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0446, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.23979637026786804, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0454, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.3095552921295166, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0449, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.2585015594959259, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0473, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.18226519227027893, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0472, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2568792402744293, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0443, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.27914026379585266, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0423, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.21452564001083374, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0457, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.22575126588344574, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0458, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.24053922295570374, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0453, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.21956735849380493, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.048, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2325354665517807, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0439, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.1692194789648056, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.044, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.25796884298324585, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0416, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.2719244956970215, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0445, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.23312833905220032, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0488, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.24458816647529602, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0473, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.24380528926849365, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0411, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.1836114227771759, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0433, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.21051307022571564, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0462, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.18093563616275787, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.045, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2444952130317688, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0485, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.19429947435855865, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0487, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.2294744998216629, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0479, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.2147439569234848, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0492, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.2777695059776306, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0419, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22944948077201843, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0458, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.2822294235229492, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0429, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.20201396942138672, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0435, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.1961182802915573, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0441, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.1653209924697876, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0473, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.2082156240940094, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0508, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.199684277176857, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0466, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.20871016383171082, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0436, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.22781118750572205, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0504, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2652275562286377, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0467, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.20812813937664032, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0456, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2533996105194092, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0458, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.18139654397964478, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0441, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.27243369817733765, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0472, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.326709121465683, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0487, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.26698657870292664, + "learning_rate": 2.999617967180403e-05, + "loss": 0.05, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.4119804799556732, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0455, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.21175070106983185, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0462, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.22503416240215302, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.0468, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.254734605550766, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0471, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26713427901268005, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0439, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.2739362120628357, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0455, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.26221534609794617, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0451, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26752719283103943, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.23052464425563812, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0481, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.162479966878891, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0435, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.24716691672801971, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0459, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.14981329441070557, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0407, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.23326952755451202, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0459, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.2200203239917755, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.046, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.21501530706882477, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0433, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.22950918972492218, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0454, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.2042679488658905, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0444, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.16468611359596252, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0434, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.2180141657590866, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0447, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.1658746749162674, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.24921023845672607, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0398, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0445, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.25189560651779175, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0429, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.2614912688732147, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0427, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.2890806794166565, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0458, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.24215207993984222, + "learning_rate": 2.999336925698801e-05, + "loss": 0.04, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.1806204915046692, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0435, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.18884161114692688, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0422, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.2760990262031555, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0427, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.22660709917545319, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0459, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.24862384796142578, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.045, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.1911306381225586, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0433, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.2612020969390869, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0414, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.20945195853710175, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.043, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.2436099350452423, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0472, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.23788926005363464, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0455, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.20128341019153595, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0456, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.24002164602279663, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0414, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.2653037905693054, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.042, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.15122903883457184, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0395, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.19299687445163727, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0452, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.2534617781639099, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0445, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.29057401418685913, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0398, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.19744732975959778, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.041, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.25267109274864197, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0439, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2406993955373764, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0445, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.2561663091182709, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0465, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.2137032002210617, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0407, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.3002893030643463, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0433, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.32190272212028503, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0444, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.2076684534549713, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0432, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.31036099791526794, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0431, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.21386218070983887, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0442, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.19012001156806946, + "learning_rate": 2.998948184763285e-05, + "loss": 0.048, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.2310018688440323, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0466, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.1984279453754425, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0443, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.24101053178310394, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0446, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.17791084945201874, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0403, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.2016495168209076, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0404, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.19435155391693115, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.21197110414505005, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0398, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.22241441905498505, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.20825718343257904, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0434, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.24533888697624207, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.046, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.2455352544784546, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0412, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.18937505781650543, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0425, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.22356615960597992, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0418, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.23797748982906342, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0457, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.23130296170711517, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0427, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.20548853278160095, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0439, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19323743879795074, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0432, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.24339382350444794, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0412, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.18780124187469482, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0416, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.19353246688842773, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0461, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.21522872149944305, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0412, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.26847484707832336, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.3781932294368744, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0401, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.3219687342643738, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0423, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.1699237823486328, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0421, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.2174658179283142, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0398, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.2433294802904129, + "learning_rate": 2.998488804893431e-05, + "loss": 0.042, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.2116415947675705, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0405, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.22813065350055695, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0432, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.24252736568450928, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.048, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.2009168565273285, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0431, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.19379755854606628, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.23313777148723602, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0384, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.22056345641613007, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0459, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.3403800129890442, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0439, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.17649835348129272, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0419, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.20594412088394165, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0418, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.2901265323162079, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0405, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.22897225618362427, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0424, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.20961393415927887, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0438, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.28585514426231384, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0477, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2239968478679657, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0398, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.1651894897222519, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0405, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.1731943041086197, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0407, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.1708967089653015, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0417, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.17396977543830872, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0359, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.21327704191207886, + "learning_rate": 2.998095036004704e-05, + "loss": 0.045, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.15585461258888245, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0403, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.21774765849113464, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0403, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.1889292299747467, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0408, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.1639491617679596, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0425, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.23745876550674438, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.043, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19481611251831055, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0453, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.21001650393009186, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0429, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.2114192098379135, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0419, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.2506379783153534, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0392, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.21439281105995178, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0422, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.24763159453868866, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0433, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.1719428300857544, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0422, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.25943970680236816, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0425, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22226977348327637, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0448, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.2018195390701294, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0408, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.20869500935077667, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0443, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.23311494290828705, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.042, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.2501373589038849, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0382, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.2650911509990692, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0377, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.21386376023292542, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0405, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.27908873558044434, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0431, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.18731699883937836, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0402, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.2566622197628021, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0419, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.2418505847454071, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0406, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.199744313955307, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0424, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.20134931802749634, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.042, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.15607263147830963, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0453, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.22614715993404388, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0413, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.22810052335262299, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0433, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.16557516157627106, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0424, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.20254060626029968, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0394, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2068638950586319, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0408, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.27261367440223694, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0433, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.23216570913791656, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0445, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.2546656131744385, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0381, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.21829853951931, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0472, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.17109332978725433, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0396, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.1746513843536377, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0413, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.22411789000034332, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0373, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.21182285249233246, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.038, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.19430142641067505, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0419, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.22196920216083527, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0395, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.20695635676383972, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0407, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.1721169799566269, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0396, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.23676829040050507, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0396, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.3019765317440033, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0443, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.23958736658096313, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0402, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.17959944903850555, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0434, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.1563226878643036, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0397, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.16746941208839417, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0402, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.26673364639282227, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0411, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.23634585738182068, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0427, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.19724313914775848, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0408, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.2072090059518814, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0389, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.18574072420597076, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0427, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.23390066623687744, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0411, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.2137259542942047, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0395, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.21458280086517334, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0422, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.1838904619216919, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.043, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.1834891140460968, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0389, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.2012249082326889, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.27410435676574707, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0473, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.21607685089111328, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0405, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.17525449395179749, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.043, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.23137639462947845, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0414, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.20958152413368225, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0404, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.21453802287578583, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.043, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.2522047758102417, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.04, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.23595386743545532, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0396, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.23375281691551208, + "learning_rate": 2.996358559911323e-05, + "loss": 0.042, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.21628527343273163, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0434, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.2463727742433548, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0391, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.19919127225875854, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.043, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.28369203209877014, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0463, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.30222833156585693, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0397, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.19382458925247192, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0402, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2075910121202469, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0392, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.2503066658973694, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0427, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.18662676215171814, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0397, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.2432677149772644, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0425, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.19900213181972504, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0422, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.12742096185684204, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.20504114031791687, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0404, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.19127227365970612, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0425, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.18699967861175537, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0419, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.20074906945228577, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0381, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.24741096794605255, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0389, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.22948721051216125, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0435, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.17772004008293152, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0412, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.228739395737648, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0406, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.18312685191631317, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0401, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.21340692043304443, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0403, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.17110474407672882, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0388, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.19640640914440155, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.2386539727449417, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.043, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.20800922811031342, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0412, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1889747679233551, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0377, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.1801699548959732, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0402, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.20975525677204132, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0426, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.22626695036888123, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0404, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.2192756086587906, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0438, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.21307197213172913, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0447, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.28529244661331177, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0412, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.26367515325546265, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.044, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.29029133915901184, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.14019586145877838, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0427, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19176369905471802, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0408, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.19753597676753998, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0395, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1900634765625, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0427, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.21698379516601562, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0452, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.21722441911697388, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0417, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.1760135143995285, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0405, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18030588328838348, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0425, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.20307210087776184, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0416, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.19713939726352692, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0401, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.18719582259655, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.19560874998569489, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.04, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1633717119693756, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0424, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.20183435082435608, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0397, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.23784901201725006, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0438, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.17358708381652832, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0393, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18991878628730774, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0387, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.20455898344516754, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.042, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.23264537751674652, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0447, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2586922347545624, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0402, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.20447152853012085, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0406, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.1814388930797577, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0412, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2679784595966339, + "learning_rate": 2.994497917870469e-05, + "loss": 0.039, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.1818714141845703, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0423, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.2153957635164261, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0391, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.1858970820903778, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0394, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.23248426616191864, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.17587599158287048, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0367, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.18720309436321259, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.039, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.17602671682834625, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0393, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.15901489555835724, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0418, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.18207870423793793, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0401, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.22800953686237335, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0416, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.20031091570854187, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0417, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.24054989218711853, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0438, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.201461061835289, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0369, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.15697605907917023, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0397, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.16894802451133728, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0403, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2567543089389801, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0421, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.20297718048095703, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0396, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.17331846058368683, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.038, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.2214743047952652, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0405, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.2107590287923813, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0388, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.1768934726715088, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0364, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.16018562018871307, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0405, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.18402984738349915, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0425, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.21482129395008087, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0399, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.20554907619953156, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0388, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.2401154637336731, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0369, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.15072989463806152, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0359, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.1649719625711441, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0399, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.2386736422777176, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0403, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.2389005720615387, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0416, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.14930105209350586, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0372, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.18874907493591309, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0356, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.21466751396656036, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0391, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1697709560394287, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.24686385691165924, + "learning_rate": 2.993190369330068e-05, + "loss": 0.04, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.19733406603336334, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0417, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.3981468677520752, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0375, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.20926636457443237, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0384, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.18989330530166626, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0365, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17171184718608856, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0386, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.20474061369895935, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0413, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.19314105808734894, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0358, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.1794378012418747, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0409, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.17263059318065643, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0385, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.17863675951957703, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0355, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.20071984827518463, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0386, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.2025180608034134, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0409, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.21840184926986694, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0397, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.197452113032341, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0379, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.15370187163352966, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0379, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.22221392393112183, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0396, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.18536874651908875, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.18744619190692902, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0378, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.21197383105754852, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0399, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.29221004247665405, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0355, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.2411249279975891, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.039, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.3187369108200073, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.036, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.18096214532852173, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0372, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.2743191719055176, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0401, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.1864418387413025, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0384, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.15518276393413544, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.039, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.18089881539344788, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0359, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.2218618243932724, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0427, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.17051172256469727, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0401, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.15884852409362793, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0378, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1561235636472702, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0383, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.19711007177829742, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0409, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.18595683574676514, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.043, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.18613699078559875, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0357, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.1390007734298706, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0356, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.25603482127189636, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0377, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.1902502328157425, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0381, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.18781402707099915, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0393, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.20504912734031677, + "learning_rate": 2.991569760600352e-05, + "loss": 0.042, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.17359918355941772, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0373, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.2029414027929306, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0412, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.23485742509365082, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0398, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1454097479581833, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0394, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.22343166172504425, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0386, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.17754614353179932, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0358, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.14276635646820068, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0405, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.2062550187110901, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0429, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2093406766653061, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.038, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.14510402083396912, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0374, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.22726939618587494, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0373, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.14905233681201935, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0392, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.15828309953212738, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0361, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.990945968675917e-05, + "loss": 0.037, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.1867590844631195, + "learning_rate": 2.990900562792633e-05, + "loss": 0.038, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.18278446793556213, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0358, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.24628379940986633, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0393, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.2195759266614914, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0372, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.2397889941930771, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0352, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.1563243269920349, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0361, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.16278347373008728, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0366, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1776551604270935, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0369, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.19893717765808105, + "learning_rate": 2.990533239951709e-05, + "loss": 0.039, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.21403104066848755, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0378, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.19885864853858948, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0393, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.18425336480140686, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0404, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.1930677592754364, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0369, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18271560966968536, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0351, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.16518554091453552, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0367, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.17990152537822723, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0349, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.19625109434127808, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0368, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.1891077160835266, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0407, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.20162540674209595, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.17586444318294525, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0346, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.13146457076072693, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0338, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.16059468686580658, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0366, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.1724517047405243, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0395, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.1847251057624817, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0365, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.20104378461837769, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.037, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.2712373733520508, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0347, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.18014875054359436, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0388, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.17479945719242096, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0388, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.24176514148712158, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0387, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.2453809529542923, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0381, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2105131894350052, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0369, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1763688176870346, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0384, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.16223029792308807, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0348, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.173785999417305, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.038, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.23269419372081757, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0393, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.19176329672336578, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0399, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.1559731364250183, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0375, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.15701691806316376, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.22872254252433777, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0371, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15535064041614532, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0378, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.21613062918186188, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0382, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.1504523903131485, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0386, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.177171990275383, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2307550013065338, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0354, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.2215310037136078, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0371, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.23132716119289398, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0389, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.1937142014503479, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0373, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.23848143219947815, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0355, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.23107734322547913, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0359, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1455046832561493, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0404, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.21890339255332947, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0421, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.1897345334291458, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0357, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.2016463577747345, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0363, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.15850050747394562, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0391, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.18169008195400238, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0377, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.26095741987228394, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0391, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.20983633399009705, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0355, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.21026623249053955, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.2104070633649826, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0343, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1801508367061615, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0369, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.22245049476623535, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.0358, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.19220919907093048, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0354, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.2285826951265335, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0391, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.15382888913154602, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0374, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.13551367819309235, + "learning_rate": 2.987759237035982e-05, + "loss": 0.039, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.23362696170806885, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0367, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.15972478687763214, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0319, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.23858226835727692, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0354, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.17988596856594086, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0391, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.21262454986572266, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0392, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.1443355828523636, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0384, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.24133095145225525, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0382, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.175334632396698, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.039, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.20340824127197266, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0376, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.16549751162528992, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0371, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.16527467966079712, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0373, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1418737769126892, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0378, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18754976987838745, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0355, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.13649867475032806, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0329, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.21408453583717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0348, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.20929333567619324, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.24083013832569122, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0378, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17924576997756958, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0369, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.22360439598560333, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0382, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.15963956713676453, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0361, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.19003787636756897, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0365, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2458433210849762, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0375, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.22048939764499664, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0382, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.19965334236621857, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0371, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.228081613779068, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0375, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.24077889323234558, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0347, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.17692111432552338, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0365, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.1996857225894928, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0358, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.24799935519695282, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0377, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.24300269782543182, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0362, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.2230309098958969, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0364, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.16640354692935944, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0346, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.14670157432556152, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.039, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18531887233257294, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0353, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.19566190242767334, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0391, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.1967019885778427, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0355, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.2161232978105545, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0375, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.17630350589752197, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0397, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18905608355998993, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0347, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.1890578716993332, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.036, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.15397298336029053, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0401, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0362, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.20983757078647614, + "learning_rate": 2.985388676461079e-05, + "loss": 0.036, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.1554982215166092, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0378, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.1961488276720047, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0356, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.17265456914901733, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0357, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.1275556981563568, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0353, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.1667783409357071, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0354, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.27920690178871155, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0385, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.12451875954866409, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0352, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.15794622898101807, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0339, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.16285695135593414, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0369, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.18990163505077362, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0376, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.1867877095937729, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0363, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.15270273387432098, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0394, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.1934751272201538, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0382, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16943520307540894, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0369, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.20546840131282806, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0336, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.18490619957447052, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0415, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.16439616680145264, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0384, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.13623777031898499, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0379, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.18209043145179749, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0357, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.19922122359275818, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0365, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.15637168288230896, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0323, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1781233251094818, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0329, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.17547845840454102, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0344, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.1770639270544052, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0358, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.2677673101425171, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0356, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.16740617156028748, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0344, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.13076575100421906, + "learning_rate": 2.983793562696423e-05, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.17586535215377808, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0381, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.1623464673757553, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.037, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.16550037264823914, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0373, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.20314498245716095, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0366, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.19387134909629822, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0377, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.17531387507915497, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0381, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.21466659009456635, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.037, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1693795919418335, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0339, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.21929262578487396, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0332, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.2253660261631012, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0339, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.2084304839372635, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0363, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.2513981759548187, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0357, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.19585898518562317, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0353, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.22262930870056152, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0357, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.19813011586666107, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0347, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.16626323759555817, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0346, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.271424800157547, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.036, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2283792346715927, + "learning_rate": 2.982684510938631e-05, + "loss": 0.04, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.15996390581130981, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0394, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.20269536972045898, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0387, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.20575514435768127, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0386, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.15178649127483368, + "learning_rate": 2.982433099619666e-05, + "loss": 0.037, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.23187878727912903, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0355, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.12988153100013733, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0369, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19367633759975433, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0367, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.15319304168224335, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0337, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.19311629235744476, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0383, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.14771778881549835, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0393, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.20381954312324524, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0414, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22292324900627136, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0358, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.13758333027362823, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0359, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.18837112188339233, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0339, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.1757460981607437, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0352, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.24256370961666107, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0346, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.2417164295911789, + "learning_rate": 2.981603574229741e-05, + "loss": 0.037, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.21129392087459564, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0371, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16878797113895416, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0367, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.17526096105575562, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0353, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.1812993884086609, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.037, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.16877815127372742, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0353, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.20823006331920624, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0384, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1536571979522705, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0356, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.1477498710155487, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0346, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.12916848063468933, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0318, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.15418170392513275, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0369, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.1883322149515152, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0369, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.1569332480430603, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0374, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.21140694618225098, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0365, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.26859667897224426, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0395, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.2182101309299469, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0382, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.18896763026714325, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0372, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.17398814857006073, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0338, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.14043423533439636, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0361, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.14995983242988586, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0363, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.15542328357696533, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0361, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.16333894431591034, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0337, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.18489807844161987, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0352, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.1794673502445221, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.037, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.1882123351097107, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0414, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.16483181715011597, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.04, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.1727719008922577, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0361, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.1968923956155777, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0389, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1838955134153366, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.034, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.23556186258792877, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0369, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.2518751323223114, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0331, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18682284653186798, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0371, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.2049458622932434, + "learning_rate": 2.979480675356209e-05, + "loss": 0.04, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.11846950650215149, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0347, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.21428292989730835, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0354, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.154558002948761, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0359, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1178995817899704, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0364, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.16926616430282593, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0373, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.1721247136592865, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0346, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.1975203901529312, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0351, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.17340779304504395, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0347, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.22424420714378357, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0384, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.1688208281993866, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.16542011499404907, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0343, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.15097306668758392, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0372, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.15530438721179962, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0345, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.16840992867946625, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.034, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.19959625601768494, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0375, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.15717248618602753, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0344, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.21473488211631775, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.04, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.18634898960590363, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0346, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.16032545268535614, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0377, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.19153323769569397, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0408, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.16018801927566528, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0334, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.1498652696609497, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0349, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.1652575582265854, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0344, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.15636402368545532, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0331, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.22963644564151764, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0359, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.16931135952472687, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0389, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.17646029591560364, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0322, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.19198739528656006, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0318, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.16826699674129486, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0389, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.1693607121706009, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.037, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.147409588098526, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0343, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.1984914094209671, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0329, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.18371893465518951, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0357, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.16279958188533783, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0377, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2122160941362381, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0367, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.20099543035030365, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0343, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.17249372601509094, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0325, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1106863021850586, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0336, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.2338913530111313, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0331, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.1545400768518448, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0321, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.17114387452602386, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0323, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.16218030452728271, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.037, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.14618389308452606, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0306, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.17804935574531555, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0348, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1512002944946289, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0318, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.14877542853355408, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.037, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.1693819761276245, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0325, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.17713536322116852, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0356, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.1719064563512802, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0358, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.19448526203632355, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0363, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.18047156929969788, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0326, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.20896470546722412, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0362, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.15999653935432434, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0318, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.17288383841514587, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0336, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.21483922004699707, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0367, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.19133612513542175, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0336, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.16010481119155884, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0377, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.1778886914253235, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0354, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.14997199177742004, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0326, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.16331784427165985, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0326, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.18247577548027039, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0367, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1838751882314682, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0338, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.17194764316082, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0365, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.1354181170463562, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0315, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.19473528861999512, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0345, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14786818623542786, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0354, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.1931883543729782, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0345, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.159915491938591, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0357, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14373569190502167, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0309, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.1468040943145752, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0339, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.1592605859041214, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0374, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.17073556780815125, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0365, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.13732466101646423, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0331, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.18718381226062775, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0337, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.179783895611763, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0383, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.15795794129371643, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0368, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.13824236392974854, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0343, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.2081076204776764, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0385, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.1407512128353119, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0338, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.1619846224784851, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0325, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.184702068567276, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0353, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.18360818922519684, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0349, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.17683662474155426, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0332, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.17690211534500122, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0364, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.18250274658203125, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0336, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.12902002036571503, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0344, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.1598006635904312, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0332, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.17691342532634735, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.037, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.2213180959224701, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0324, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.14762577414512634, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0321, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.14543993771076202, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.034, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.18970227241516113, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0377, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.15049827098846436, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0341, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.1485714316368103, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0361, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1893187314271927, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0354, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.2345002293586731, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0336, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.24119973182678223, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0325, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.18577414751052856, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0337, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.1941320151090622, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0355, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.17646151781082153, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0342, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.13801687955856323, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0351, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.15124225616455078, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0337, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19114945828914642, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.033, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.23505449295043945, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0357, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.15116260945796967, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.15656375885009766, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0322, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.18982963263988495, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0368, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.18300476670265198, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0379, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.16485247015953064, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0335, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.19399121403694153, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0353, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.22944431006908417, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0343, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.16755999624729156, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0351, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.15779320895671844, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0318, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.2113935351371765, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0343, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.14698673784732819, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0332, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19958600401878357, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0357, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20250289142131805, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0334, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.1955910474061966, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0381, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.13143742084503174, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0321, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.15840105712413788, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0333, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.20588421821594238, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0347, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.1881185919046402, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0345, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.16714425384998322, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.032, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.16526015102863312, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0316, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.1865321844816208, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0351, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.19202300906181335, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0325, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.15551044046878815, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0339, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.15684190392494202, + "learning_rate": 2.969840338753429e-05, + "loss": 0.034, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.16806578636169434, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0319, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.1554284542798996, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0348, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1836974322795868, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0375, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.20003411173820496, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0364, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.1747797578573227, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0328, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.2204974740743637, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0331, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.17183512449264526, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0381, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.14398148655891418, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.03, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.17428989708423615, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0329, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.13864420354366302, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0319, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.15052364766597748, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0357, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.21513856947422028, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0349, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19239531457424164, + "learning_rate": 2.968758727655112e-05, + "loss": 0.0377, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.22284342348575592, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0358, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.17206035554409027, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0349, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.1991468220949173, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0316, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.13542494177818298, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0329, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.14473386108875275, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0346, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.1972355991601944, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0348, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.17447993159294128, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0337, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.14419333636760712, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0355, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.19937965273857117, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0381, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.18988190591335297, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0331, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.1552465558052063, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0333, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.1769426017999649, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0344, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.17652767896652222, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0354, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1598699688911438, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0354, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.17556554079055786, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0372, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.18424467742443085, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0372, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.1591036170721054, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.038, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.1973133534193039, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0332, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.139867901802063, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0336, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.17536301910877228, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0339, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16841425001621246, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0319, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.24087557196617126, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.1419876366853714, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0341, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0315, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.14541317522525787, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0326, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.18411535024642944, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0382, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1769087016582489, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0355, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.16459472477436066, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0354, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.1770380288362503, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.034, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.16219645738601685, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0316, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.18393918871879578, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0366, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.1436525285243988, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0318, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.18282611668109894, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0376, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.16712592542171478, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0367, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.13408221304416656, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0305, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.17664852738380432, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0312, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.13730520009994507, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0323, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1278112679719925, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0322, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.1742715984582901, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0339, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.15022823214530945, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0348, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.17816156148910522, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0348, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.18016202747821808, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0357, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.21158602833747864, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0331, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1747511923313141, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0301, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.16840165853500366, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0363, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.1507093459367752, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0352, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.2081453651189804, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.035, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.1459163874387741, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0331, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.17468400299549103, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0339, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.15824441611766815, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0309, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.19502659142017365, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0305, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.1502636969089508, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.032, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.13870884478092194, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0356, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1596919149160385, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0347, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.1830880492925644, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0359, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.13567551970481873, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0319, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.12907959520816803, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0327, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.17357082664966583, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.0308, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.20002517104148865, + "learning_rate": 2.963613172841619e-05, + "loss": 0.035, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.13852699100971222, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0353, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.15445692837238312, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.11687827110290527, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0319, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.15454575419425964, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0366, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.1916663944721222, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0327, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.16530223190784454, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0309, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17331530153751373, + "learning_rate": 2.962976980219581e-05, + "loss": 0.032, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.17225238680839539, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0323, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.16390499472618103, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0324, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.16459178924560547, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0315, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.1416798084974289, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0348, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.1546057015657425, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0329, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15206973254680634, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0304, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.18791860342025757, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.03, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.17669418454170227, + "learning_rate": 2.962243236871436e-05, + "loss": 0.034, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.1335674524307251, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0323, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.17903365194797516, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0327, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.17771181464195251, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0341, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1532796174287796, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0328, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.18777944147586823, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0343, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.2553282082080841, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0342, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.24115385115146637, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0347, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.18773671984672546, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0324, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.18168678879737854, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0355, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.14390279352664948, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0323, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.13260799646377563, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0336, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.16469483077526093, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0341, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.1618199497461319, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0305, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.15524879097938538, + "learning_rate": 2.960942086834339e-05, + "loss": 0.034, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.14689622819423676, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0352, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.17949312925338745, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0329, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.2376648485660553, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0326, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.1302678883075714, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0356, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.16316285729408264, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0318, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.14314667880535126, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0347, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.13162434101104736, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0335, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.13411720097064972, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0334, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.20853494107723236, + "learning_rate": 2.960094148810064e-05, + "loss": 0.036, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.15634609758853912, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0314, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.16109681129455566, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0323, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.15969422459602356, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0292, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.13126519322395325, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0338, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.13577662408351898, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0323, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13226218521595, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0307, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.12283602356910706, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0305, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16819486021995544, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0316, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.1443309634923935, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.031, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.17339679598808289, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0335, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.12366446852684021, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0319, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.18073806166648865, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0355, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.16230227053165436, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.033, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.15816937386989594, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0331, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.15911494195461273, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0296, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.19500406086444855, + "learning_rate": 2.958564529914256e-05, + "loss": 0.034, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.16553588211536407, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0348, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.1780945360660553, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0297, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16839291155338287, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0317, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.26763999462127686, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0335, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.1292608082294464, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0319, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19598962366580963, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0346, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.16073159873485565, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0317, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.2037622481584549, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0324, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.11946277320384979, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0329, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.12799547612667084, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0329, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.23773977160453796, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0327, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.19241386651992798, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0371, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.15968935191631317, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0335, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.1712830811738968, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0302, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.16464142501354218, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0346, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.2122460901737213, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0319, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.16571341454982758, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.033, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1610543429851532, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0347, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14235250651836395, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0322, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.14999081194400787, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0299, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.033, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.1448807269334793, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0307, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.20123368501663208, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0315, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.16191548109054565, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.031, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.1475522518157959, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.17617955803871155, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0323, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.20633190870285034, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0287, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.13562145829200745, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0347, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.15800844132900238, + "learning_rate": 2.955719840324779e-05, + "loss": 0.034, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.1652759164571762, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0291, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.11819476634263992, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0309, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.17562547326087952, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0327, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.1934351772069931, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0342, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.1565367877483368, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.033, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.21048930287361145, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0357, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.16762706637382507, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0341, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.1267809122800827, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0322, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.18461033701896667, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0357, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.13451576232910156, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0327, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.15478897094726562, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0311, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1461619883775711, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0314, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.1432616114616394, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0339, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.14061163365840912, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0337, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.1950581669807434, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0315, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.13942912220954895, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.031, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.15901322662830353, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0324, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.19862298667430878, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0322, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.15155215561389923, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0302, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16394968330860138, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0299, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.13225089013576508, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0305, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.18203529715538025, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0319, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.15323573350906372, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0346, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.18822363018989563, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.038, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1510896533727646, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0336, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.12274178862571716, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.032, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.1287696361541748, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0341, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.16619601845741272, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0321, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.12426833063364029, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0298, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.1751766800880432, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0341, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.13578996062278748, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0306, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.19475996494293213, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0314, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.13291390240192413, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0341, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.16193236410617828, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0325, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.16108311712741852, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0356, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.16233862936496735, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0303, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.14042457938194275, + "learning_rate": 2.951955450025455e-05, + "loss": 0.031, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.14922280609607697, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0336, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.1389254927635193, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0324, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.21042059361934662, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0331, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.18529878556728363, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0311, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1345708668231964, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0316, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.14794935286045074, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0367, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.18609392642974854, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0326, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.1762595772743225, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0312, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.15549533069133759, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0342, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.1910371482372284, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0316, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.12789981067180634, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0286, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.11457879841327667, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0303, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.12534750998020172, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0313, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.17089471220970154, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0318, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.16774669289588928, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0326, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.18693485856056213, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0356, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.20419354736804962, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15851624310016632, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0328, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1284184604883194, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.1148369237780571, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0291, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.15302640199661255, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.037, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.16082480549812317, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.031, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.16006684303283691, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0346, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14866025745868683, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0323, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.15490789711475372, + "learning_rate": 2.949326462724779e-05, + "loss": 0.033, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.18960629403591156, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0302, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.16329076886177063, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0338, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1700190156698227, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0326, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.14541998505592346, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0332, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.13809162378311157, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.16794940829277039, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0345, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1897762268781662, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0304, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2064688503742218, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0332, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.14193041622638702, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0332, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.15571711957454681, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0339, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.14264710247516632, + "learning_rate": 2.948147907679037e-05, + "loss": 0.036, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.1411963850259781, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0344, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.18891720473766327, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0308, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.15993206202983856, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0339, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.15609434247016907, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0315, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.19272638857364655, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0325, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.2836095690727234, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0323, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16382543742656708, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0335, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.18899215757846832, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0301, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.17764721810817719, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0322, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.1808568239212036, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0304, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.166172593832016, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0298, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.18262767791748047, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0305, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.17854458093643188, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.033, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.16440987586975098, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.033, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.19478020071983337, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0288, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.2605016231536865, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0318, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1850014477968216, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0313, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.1339086890220642, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.17828164994716644, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0315, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.1235261932015419, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0328, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.13704872131347656, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0322, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1296597719192505, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0315, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.16906750202178955, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0347, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14643864333629608, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.031, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.16102853417396545, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0315, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.17575663328170776, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0312, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1488381028175354, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0318, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.14629733562469482, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0333, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.17051763832569122, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.034, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.14636048674583435, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0315, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.19957765936851501, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.032, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.15027561783790588, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0342, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1580146700143814, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.032, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1452188491821289, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0302, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.1924576759338379, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0323, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.199595108628273, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0326, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.1557701975107193, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0303, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.21232250332832336, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0324, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.13840343058109283, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0321, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.14219941198825836, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0314, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1747855842113495, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0342, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1730964183807373, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0334, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.12938249111175537, + "learning_rate": 2.94341324778745e-05, + "loss": 0.029, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1782408058643341, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0299, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.943188093053675e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.12131186574697495, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0277, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.19141671061515808, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0307, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.2431664764881134, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0326, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15145368874073029, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0329, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.19781085848808289, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0332, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.19477561116218567, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0317, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.1818031370639801, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0352, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.13725785911083221, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0272, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.1874770075082779, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0322, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.18701452016830444, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0304, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.15173715353012085, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.032, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.1315767467021942, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0331, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.16627946496009827, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0299, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.2122701108455658, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0308, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.12093837559223175, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0303, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.2072226107120514, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0331, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.14395366609096527, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0323, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.18454021215438843, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0349, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1647966057062149, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0342, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1907777190208435, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0347, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.14438189566135406, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0295, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.16802345216274261, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0283, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.1488419622182846, + "learning_rate": 2.940567320758981e-05, + "loss": 0.033, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14881044626235962, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0318, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.18762606382369995, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0322, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.15426765382289886, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0314, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.17882433533668518, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0361, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.16833186149597168, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0319, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.16899487376213074, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0283, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.15301142632961273, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0325, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.13270439207553864, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0305, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.16603541374206543, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0347, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.21959784626960754, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0299, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.24568438529968262, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0339, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.14889934659004211, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0284, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.20354624092578888, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0309, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.18124797940254211, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0352, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.13871392607688904, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0311, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.18997471034526825, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0299, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.09903991222381592, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0304, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.13336266577243805, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0306, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.1564595103263855, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0295, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.17495550215244293, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.17048844695091248, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0266, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.1257491409778595, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0299, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.15244479477405548, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0341, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.15737923979759216, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.03, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.13226579129695892, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0268, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.21549297869205475, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0323, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.2102872133255005, + "learning_rate": 2.937416924082395e-05, + "loss": 0.029, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.14369380474090576, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0306, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.1390914022922516, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0273, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.13536471128463745, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0299, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.13321688771247864, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0298, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13358351588249207, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0309, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.1732599437236786, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0319, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.17455634474754333, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0314, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.1714484691619873, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0345, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.18120428919792175, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0293, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.21681863069534302, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0335, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.12099924683570862, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0299, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.13233523070812225, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0372, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.16708864271640778, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0305, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.2122814804315567, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0319, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14159013330936432, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0343, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.13396364450454712, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0308, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.17630934715270996, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0349, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.16417403519153595, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0316, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.16593796014785767, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0291, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.13613900542259216, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0299, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.15980054438114166, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0319, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1578991562128067, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0288, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.1628173291683197, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0317, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.15937024354934692, + "learning_rate": 2.934549758921277e-05, + "loss": 0.031, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.15258124470710754, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0323, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.16142596304416656, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0289, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.15643659234046936, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0295, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.17128989100456238, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0313, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.1871197372674942, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0314, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.1448175013065338, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0318, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.21338209509849548, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0334, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15262271463871002, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0293, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.14302000403404236, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0307, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.13756762444972992, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0288, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.1390128880739212, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0286, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.16088609397411346, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0299, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.11852050572633743, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0284, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.17093200981616974, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0331, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.2325763702392578, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0269, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.19948336482048035, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0314, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.16918544471263885, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0311, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.14563684165477753, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0301, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.19645430147647858, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0312, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.18950732052326202, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0292, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.11659497767686844, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0307, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1314595341682434, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0326, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1457567662000656, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.16525784134864807, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0322, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.15977199375629425, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0334, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.17081087827682495, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0308, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.20844970643520355, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.029, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.22166086733341217, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1766587346792221, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0298, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1307239532470703, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0298, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0289, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.17264464497566223, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0317, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.15076857805252075, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0308, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14443998038768768, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0291, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.13115015625953674, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0329, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1444655954837799, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0294, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.16740760207176208, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0318, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.14201150834560394, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0314, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.14225733280181885, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0297, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.16758595407009125, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0307, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.16751281917095184, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.03, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.1855877935886383, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0333, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.17420977354049683, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0277, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.1568649560213089, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0312, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.15874658524990082, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0297, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.13727585971355438, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0334, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.14999760687351227, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0328, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.16789689660072327, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.14890174567699432, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0312, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.18001733720302582, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0322, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.1718987226486206, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.03, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.18727312982082367, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0297, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.1528918594121933, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0296, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.13372628390789032, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0288, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.14980314671993256, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0319, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.11368083208799362, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0312, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.14968374371528625, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0311, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.15593667328357697, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0287, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.1762111783027649, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0296, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.15493011474609375, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0272, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.14142747223377228, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.03, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.1307222545146942, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0294, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.1598598062992096, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.03, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.17966923117637634, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0337, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15533174574375153, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0276, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.1223452091217041, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0294, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.12733329832553864, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0289, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.15903618931770325, + "learning_rate": 2.926085774098786e-05, + "loss": 0.029, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1539735347032547, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0272, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.20487286150455475, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0312, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.19920656085014343, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0315, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.15179695188999176, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0305, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.16460831463336945, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0315, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.21031928062438965, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0283, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.13835182785987854, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0287, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.1597253829240799, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0336, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.11488606780767441, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0328, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.18707934021949768, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0328, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.18874959647655487, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0317, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.1182878389954567, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0316, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.11695742607116699, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0308, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.1597018986940384, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0324, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.17729975283145905, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0316, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.18250048160552979, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0309, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.20568659901618958, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0293, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.12907297909259796, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0281, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.1809365600347519, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0333, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.17379151284694672, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0327, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.16082972288131714, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0308, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.17415443062782288, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0299, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.1659587323665619, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0272, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.1538572907447815, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0309, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.16303570568561554, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0322, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1444631963968277, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0306, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.24458004534244537, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0285, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.15384158492088318, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0277, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1425933837890625, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0303, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.12485234439373016, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0279, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.14287105202674866, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0297, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.1621903032064438, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0268, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.145540252327919, + "learning_rate": 2.921797581830774e-05, + "loss": 0.03, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.1410767287015915, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0309, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.10953503102064133, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0299, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.15029482543468475, + "learning_rate": 2.92140191009304e-05, + "loss": 0.031, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.13262124359607697, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0299, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1503380835056305, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0329, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.18932904303073883, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.028, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.1473148614168167, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0315, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.18515919148921967, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0309, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.16122235357761383, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0279, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.1684419959783554, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0309, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.21400189399719238, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0304, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.12552157044410706, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0286, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.15383073687553406, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.029, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.1269458532333374, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0289, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.1839684545993805, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0316, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.13441112637519836, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0281, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.13045959174633026, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0296, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.13933542370796204, + "learning_rate": 2.919408984114327e-05, + "loss": 0.029, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.15143601596355438, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0296, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.16731251776218414, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.032, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.13245129585266113, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0268, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.17244738340377808, + "learning_rate": 2.918873439576597e-05, + "loss": 0.028, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.1709858626127243, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0293, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18727441132068634, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0302, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1482084095478058, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0297, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.1343366652727127, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0303, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.1456393450498581, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0299, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1569104641675949, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0281, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.19711613655090332, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0298, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.14027468860149384, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.031, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.16852174699306488, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0289, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.18446287512779236, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0308, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.16207733750343323, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0283, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.19747264683246613, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0301, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.15206509828567505, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0295, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.18873688578605652, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0286, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.144051656126976, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0335, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.14620175957679749, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0284, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.19520114362239838, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0304, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.16127106547355652, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0333, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.11467311531305313, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0303, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11777719110250473, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0311, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.12151405215263367, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0278, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1852089911699295, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0297, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.19555963575839996, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0321, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.1584533303976059, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0292, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.1764054298400879, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0306, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.16257385909557343, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0298, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.13896049559116364, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0296, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.16509538888931274, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0321, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.14912301301956177, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0285, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.18058481812477112, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0302, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.18734610080718994, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0289, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.19707795977592468, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0327, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.16215947270393372, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0302, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1710345447063446, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0332, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.19638675451278687, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0316, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.17807847261428833, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0285, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.13119569420814514, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0298, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.13414712250232697, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0301, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.1941734254360199, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0327, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.19624412059783936, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0291, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.17615875601768494, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0293, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.19011878967285156, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0316, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.15080299973487854, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0263, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.1676572859287262, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0318, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.14975300431251526, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0263, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17691564559936523, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.17775340378284454, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0298, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.14438258111476898, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.14729741215705872, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0289, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.20924119651317596, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0297, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.15471352636814117, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0289, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.1571032702922821, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0307, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15048687160015106, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0301, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.15063059329986572, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0279, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15528838336467743, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0299, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.1756942719221115, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0291, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.1491880565881729, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0296, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.1562451720237732, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0302, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.12366753816604614, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0257, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.15308785438537598, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.16614912450313568, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0301, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.12428635358810425, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0303, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.135500967502594, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0283, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.16944879293441772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0279, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.174363911151886, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0295, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.16406473517417908, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0284, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.2144644856452942, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0277, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.22006124258041382, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0264, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.15261462330818176, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0298, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.15205760300159454, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.14101643860340118, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0305, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16015839576721191, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.17772945761680603, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0312, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16818812489509583, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0283, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1448637843132019, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0325, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.14702047407627106, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0303, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.15547925233840942, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0299, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.18527188897132874, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.029, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.1533391922712326, + "learning_rate": 2.907944011202334e-05, + "loss": 0.029, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.15355326235294342, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0286, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.1503538340330124, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0312, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.12572433054447174, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0277, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.13614071905612946, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0303, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.13363786041736603, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0292, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.13318102061748505, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.0289, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.14687903225421906, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0301, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.13144652545452118, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0271, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.16351011395454407, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0292, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.19256997108459473, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0268, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13006120920181274, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0285, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.12932874262332916, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.11981740593910217, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.14302268624305725, + "learning_rate": 2.905937392537971e-05, + "loss": 0.029, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.1446913331747055, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0289, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.1500432789325714, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.15604284405708313, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0283, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.1795264482498169, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0289, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.14356553554534912, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0304, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.1323058009147644, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0302, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.2098175585269928, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0297, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.1424631029367447, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0295, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.1956447958946228, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0294, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.11993541568517685, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0273, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.18414588272571564, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0316, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.17320625483989716, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0319, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.17650462687015533, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0304, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.18740026652812958, + "learning_rate": 2.903909846652783e-05, + "loss": 0.034, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.12898528575897217, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0309, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.11721221357584, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0297, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.1659565567970276, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0294, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.1469241976737976, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0298, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14588896930217743, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0332, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.16243143379688263, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0305, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.15857404470443726, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0318, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.14992372691631317, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0305, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.13760966062545776, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0299, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.14302772283554077, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0313, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.18560320138931274, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0262, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.13117150962352753, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0286, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.16370882093906403, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0298, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.18689852952957153, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.031, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.16412676870822906, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0289, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.20153313875198364, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0297, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.15255704522132874, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0316, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.2036312222480774, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0306, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.16087041795253754, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0315, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.1514595001935959, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0295, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.13961558043956757, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0259, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1996241807937622, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0336, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.15389376878738403, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0278, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.12345903366804123, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.032, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.16360828280448914, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0292, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.19773486256599426, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.032, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.11931311339139938, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0268, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.1378420740365982, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0296, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.09076640754938126, + "learning_rate": 2.899643488777395e-05, + "loss": 0.026, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.1116260215640068, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0266, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.18670536577701569, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0314, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.15351636707782745, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0279, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.15310993790626526, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0298, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.15458840131759644, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0329, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.12034284323453903, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0266, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.14272604882717133, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0296, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.16335424780845642, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0266, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.16105882823467255, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0355, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.1455371230840683, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0261, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.21555180847644806, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0301, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.15278705954551697, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0292, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.14099596440792084, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0287, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.10848120599985123, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.028, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.1380046308040619, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0273, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.14006251096725464, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0328, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.16518674790859222, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0286, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.13295377790927887, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.031, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.1623251736164093, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0281, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.16029980778694153, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.03, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.15575948357582092, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0294, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.157670795917511, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0294, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1470523178577423, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0281, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.12192463129758835, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0248, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.1240374892950058, + "learning_rate": 2.895893835994484e-05, + "loss": 0.028, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1385754942893982, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0287, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.140727236866951, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0281, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.15758906304836273, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0294, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.19664499163627625, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0277, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.12567110359668732, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0279, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.13595536351203918, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0313, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.14262230694293976, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0317, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1474338322877884, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0282, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.18601401150226593, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0274, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.1392786204814911, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0269, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.12022347748279572, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0262, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.17340584099292755, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0277, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.1438349187374115, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0273, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.1408676654100418, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0277, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.14853249490261078, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0291, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.20401965081691742, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.031, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.18385861814022064, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.12695786356925964, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0267, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.14115597307682037, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0291, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.15790492296218872, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0294, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.15104074776172638, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0278, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13686735928058624, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0276, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.12029985338449478, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0305, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1351509988307953, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0293, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.10609307885169983, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0285, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15056312084197998, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0292, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.17062142491340637, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0281, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.15565407276153564, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0298, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.136429563164711, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0282, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.1649797409772873, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0267, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.19815988838672638, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.032, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15660062432289124, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.03, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.1840171217918396, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0342, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.14063017070293427, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0282, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.14935986697673798, + "learning_rate": 2.890533053632388e-05, + "loss": 0.029, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.14691074192523956, + "learning_rate": 2.89037798520224e-05, + "loss": 0.033, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.14992092549800873, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0262, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.16885656118392944, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0301, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1223178282380104, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0275, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.1200852021574974, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0277, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.1251501888036728, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0252, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.16905754804611206, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0277, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.13312722742557526, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0296, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.13198228180408478, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0269, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.14326612651348114, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0253, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.14861920475959778, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0262, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.19368910789489746, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0295, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.143910750746727, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0282, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.12314137071371078, + "learning_rate": 2.888352491214935e-05, + "loss": 0.027, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.16095006465911865, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0272, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.1414606273174286, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.029, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.15631091594696045, + "learning_rate": 2.887882538107974e-05, + "loss": 0.0291, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16476406157016754, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0283, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.2236901968717575, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0286, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.1318763792514801, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0274, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.15661245584487915, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.029, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.13406990468502045, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0287, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16861270368099213, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0272, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.18082283437252045, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0285, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.17517460882663727, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0277, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.17011325061321259, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0325, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14378181099891663, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0306, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1766553819179535, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0304, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.1666528880596161, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0294, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.15178363025188446, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0273, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.14275510609149933, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0291, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.1568918377161026, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0267, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.1370164304971695, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0268, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.10388749092817307, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0289, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11448585987091064, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0302, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.15148931741714478, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0292, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.14388935267925262, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0257, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.14465516805648804, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0272, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.13414210081100464, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0273, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12774589657783508, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.028, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.24874050915241241, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0282, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.1624344289302826, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0294, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.12817884981632233, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0324, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1509743630886078, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0269, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.31345340609550476, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0287, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.1523321270942688, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0259, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.1633133739233017, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0269, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.14278745651245117, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.0253, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.18361203372478485, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0282, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.10244551301002502, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0268, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.125647634267807, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0264, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.12238023430109024, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0264, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.16535502672195435, + "learning_rate": 2.882169202177678e-05, + "loss": 0.0329, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.19624854624271393, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0303, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.18285983800888062, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0266, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.13460052013397217, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.0293, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.1427820920944214, + "learning_rate": 2.881525979398615e-05, + "loss": 0.0244, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.14855900406837463, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0276, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.1433439403772354, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0299, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.1452850103378296, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0271, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.14403793215751648, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0264, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.13841332495212555, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0277, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.1426801085472107, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0296, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.14492706954479218, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0272, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.1315297782421112, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.0289, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.1513436734676361, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0284, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.15784604847431183, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.0273, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.12729156017303467, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0297, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.20208381116390228, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0284, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.1486770659685135, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0271, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.1463368982076645, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.0292, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.12362723797559738, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0252, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.18984590470790863, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0292, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.12962038815021515, + "learning_rate": 2.878773564250086e-05, + "loss": 0.0301, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.132072314620018, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0298, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.1482372134923935, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0267, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.15687556564807892, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0308, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.11656040698289871, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.028, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.16482318937778473, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0256, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.13332504034042358, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.027, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.16453887522220612, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0253, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.16902299225330353, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0283, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.12483953684568405, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0276, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.14033488929271698, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0277, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.18922871351242065, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0258, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.15026360750198364, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0298, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17492659389972687, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0275, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.14793051779270172, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0301, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.17575892806053162, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.027, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.1661873608827591, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0275, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.14529843628406525, + "learning_rate": 2.875990888327069e-05, + "loss": 0.0277, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.11084084957838058, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0278, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1433780938386917, + "learning_rate": 2.875661528053842e-05, + "loss": 0.026, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.13195206224918365, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0298, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.14103750884532928, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0309, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.17067542672157288, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0296, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.14078323543071747, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0288, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.17906904220581055, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0268, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.163010373711586, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0285, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.17636623978614807, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0276, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.18726587295532227, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0297, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.18583180010318756, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.16773444414138794, + "learning_rate": 2.874008460348417e-05, + "loss": 0.0289, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.14579996466636658, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.12655268609523773, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0246, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.15124735236167908, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0297, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.16273140907287598, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0295, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14014655351638794, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0262, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12657561898231506, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0292, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.13813915848731995, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.0287, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.12874144315719604, + "learning_rate": 2.872678492608539e-05, + "loss": 0.026, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.12086840718984604, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0245, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.15290603041648865, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0297, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.13261260092258453, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0263, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.17371073365211487, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0268, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.17300288379192352, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0264, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.16441872715950012, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0279, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.17883789539337158, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.03, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.14519335329532623, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0266, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.18063998222351074, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0275, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.19818824529647827, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0256, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1695982664823532, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0274, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.16905854642391205, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0277, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.1852719783782959, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.03, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.14601513743400574, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0261, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.11720956116914749, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0279, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.14177089929580688, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0268, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.14933669567108154, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0292, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15183576941490173, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0285, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.1255270391702652, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.025, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.1467735320329666, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0258, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.13636574149131775, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0278, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.12642863392829895, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0254, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.09835807979106903, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0284, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.11784205585718155, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0285, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.14453290402889252, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0301, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.19569514691829681, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0264, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.13860125839710236, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.026, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.16651827096939087, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0283, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.13682647049427032, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0309, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.14632774889469147, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0275, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.1401532143354416, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0261, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.12643243372440338, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0268, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1152961477637291, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0259, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.14027607440948486, + "learning_rate": 2.866951776928e-05, + "loss": 0.0306, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.12379784137010574, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0272, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.15595580637454987, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0288, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.12823393940925598, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0263, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1481468379497528, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0283, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.12768998742103577, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0262, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.15771906077861786, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0267, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.18978171050548553, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0295, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.1386374831199646, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0277, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.1583528220653534, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0273, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.09999477118253708, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0253, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.14432361721992493, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0289, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.19001086056232452, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.0295, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.13931159675121307, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0257, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.1621151715517044, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0269, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.15239235758781433, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0275, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12619581818580627, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0273, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.14698320627212524, + "learning_rate": 2.864043385528954e-05, + "loss": 0.0275, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.14121954143047333, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0271, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.16203583776950836, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0265, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.15562862157821655, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0267, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.16438758373260498, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0285, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.13769859075546265, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0274, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.14908984303474426, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0256, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.14057780802249908, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.0261, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.16056638956069946, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.0267, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.12299710512161255, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0257, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.12815117835998535, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0284, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.14666494727134705, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0303, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.15829558670520782, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0243, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1499258577823639, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.029, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.16584458947181702, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0272, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1765419989824295, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0278, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.17066679894924164, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0281, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13183729350566864, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0255, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.12179457396268845, + "learning_rate": 2.86093128283797e-05, + "loss": 0.028, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.10134928673505783, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0275, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16393089294433594, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0278, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1360260546207428, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0297, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.16158054769039154, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0249, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.12037541717290878, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.0286, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.16677190363407135, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0277, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1442420333623886, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0268, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.1756538599729538, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0282, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.11580783128738403, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0282, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.11671322584152222, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0279, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.15163642168045044, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0297, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.1390339583158493, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0282, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.11784064769744873, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0275, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.15270282328128815, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0282, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.09030280262231827, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0281, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.15108083188533783, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0274, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.1403200924396515, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0265, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.14890284836292267, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0303, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.11579708009958267, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0283, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.14590834081172943, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0268, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.1513289511203766, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0264, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.13494084775447845, + "learning_rate": 2.857082133984191e-05, + "loss": 0.026, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.11738834530115128, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.029, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.1497882753610611, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0271, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.17350555956363678, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0277, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.16718043386936188, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0269, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.12658995389938354, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0267, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15880689024925232, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0256, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.1480095237493515, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0263, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.13779640197753906, + "learning_rate": 2.85567006866004e-05, + "loss": 0.0282, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1088612824678421, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0245, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.17420272529125214, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0267, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.12365500628948212, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.025, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.13261641561985016, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0256, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.10485266894102097, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0253, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14070947468280792, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0284, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.1179763451218605, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0241, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13687312602996826, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0284, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.17160922288894653, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0263, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.17487400770187378, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0275, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.11352817714214325, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1871306300163269, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0277, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.13434743881225586, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0278, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.16361933946609497, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0257, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.19457586109638214, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0277, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.16905367374420166, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0323, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.14845560491085052, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0248, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.12934044003486633, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.026, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.11046823114156723, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0267, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.14174290001392365, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0255, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.15194296836853027, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.0257, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.1532754749059677, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0269, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19864243268966675, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0288, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.16765300929546356, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0271, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.15602844953536987, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0288, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.1571321040391922, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0264, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1625281423330307, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0268, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.12808986008167267, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0307, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.12471310794353485, + "learning_rate": 2.850496147627652e-05, + "loss": 0.027, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.1521507054567337, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0289, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1389119178056717, + "learning_rate": 2.850136142292186e-05, + "loss": 0.024, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1338133066892624, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13214652240276337, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.0274, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.17002162337303162, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0277, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.1433296948671341, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0267, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.19547788798809052, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0287, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.1916888952255249, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0296, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.17188303172588348, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0262, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.14335697889328003, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.0269, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14298376441001892, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0278, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.12892583012580872, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0278, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.16246432065963745, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0255, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.1237327978014946, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0252, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.12249717861413956, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0238, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.16828545928001404, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0262, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.12834671139717102, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0287, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.11583935469388962, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.029, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.10479022562503815, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0244, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.18407097458839417, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0276, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.12898766994476318, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0277, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.13034385442733765, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0241, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.13197654485702515, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0269, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.14327986538410187, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0281, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.15228267014026642, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0269, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.11858481913805008, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0261, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.16075590252876282, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0284, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.1527116447687149, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0283, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.1280883252620697, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0294, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1348433941602707, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0251, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.15039700269699097, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0272, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.16682399809360504, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.0277, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.11187390983104706, + "learning_rate": 2.844503681416359e-05, + "loss": 0.0298, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.1675758808851242, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0267, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.16022244095802307, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0301, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.16192372143268585, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0277, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.13092738389968872, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.12416383624076843, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.028, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.12514744699001312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0252, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.1705007702112198, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.0274, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.14251261949539185, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0264, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.147093266248703, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0279, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.12747326493263245, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0265, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.14222867786884308, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0276, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.12574070692062378, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0262, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.12812504172325134, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0225, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.13585571944713593, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0267, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.12168630957603455, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0236, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.12048719823360443, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0225, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.15165774524211884, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0267, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.15395723283290863, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.026, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.14715734124183655, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0259, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.11370360106229782, + "learning_rate": 2.840817735753433e-05, + "loss": 0.0284, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1264869123697281, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.028, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.13726982474327087, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0268, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1450515240430832, + "learning_rate": 2.840261328154438e-05, + "loss": 0.0251, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.1328781247138977, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.025, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.15474216639995575, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.0292, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.1338219791650772, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0279, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.16368664801120758, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.027, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.18125605583190918, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.0252, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.127846822142601, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0252, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.17743080854415894, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0275, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.16253551840782166, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0287, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.14212867617607117, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0274, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.1380675882101059, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0263, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.12773451209068298, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.0257, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.1556098759174347, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0264, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.16690294444561005, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0296, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.11710768938064575, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0252, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.17441801726818085, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0318, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.10790088772773743, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0251, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.16537238657474518, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0266, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.1414048671722412, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0276, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.1096765398979187, + "learning_rate": 2.836716156259365e-05, + "loss": 0.0258, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.14450953900814056, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0278, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15733887255191803, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0278, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.15253902971744537, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0275, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.1410975307226181, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0249, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1242721900343895, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0251, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.15358713269233704, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0287, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.15856830775737762, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.029, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.1777517944574356, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0254, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.10823221504688263, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0262, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.144780695438385, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0286, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12900282442569733, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0243, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.15370796620845795, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0238, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.15801627933979034, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0259, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.19058606028556824, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0257, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14663439989089966, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0281, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.10609344393014908, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.0261, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.11403398960828781, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0291, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.11631747335195541, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0243, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.1325336992740631, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0266, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.11245301365852356, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.10258981585502625, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0252, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.15847361087799072, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0258, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.15267148613929749, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0246, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.15297843515872955, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0248, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.1767774075269699, + "learning_rate": 2.831995639121568e-05, + "loss": 0.0292, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.12696890532970428, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.025, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.17774799466133118, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0267, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.15603452920913696, + "learning_rate": 2.83142492407012e-05, + "loss": 0.0285, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.12731599807739258, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0272, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.13816066086292267, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0265, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.14216956496238708, + "learning_rate": 2.830853299002291e-05, + "loss": 0.026, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1427476406097412, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0263, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.14667609333992004, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0282, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.11785806715488434, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0276, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.125101238489151, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0255, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.10735481232404709, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0248, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.14256443083286285, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0286, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12700040638446808, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0275, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.11069332808256149, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0283, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.1438131332397461, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.0269, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.1901932805776596, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0274, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.17823027074337006, + "learning_rate": 2.828557706390495e-05, + "loss": 0.029, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.13579529523849487, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0267, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.1431250274181366, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0279, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.11313948035240173, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0274, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1669740527868271, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0288, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.12642791867256165, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0267, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.15632158517837524, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0256, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.1404937356710434, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0263, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.13466882705688477, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0249, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.13014933466911316, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0255, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.18688003718852997, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0267, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.14594615995883942, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0253, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.1708635836839676, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0244, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.16193392872810364, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0265, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.13249778747558594, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0286, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.13875705003738403, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0262, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.19137892127037048, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0246, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.15596376359462738, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0248, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.1577422320842743, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0247, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.14322742819786072, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0263, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.13651713728904724, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0261, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11288725584745407, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0252, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.09636299312114716, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0267, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.11729153245687485, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0275, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.14827747642993927, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0265, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.1446153074502945, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0257, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.1329014152288437, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0247, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.12726181745529175, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0259, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.13508957624435425, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0258, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.14538758993148804, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0264, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.10585593432188034, + "learning_rate": 2.822755216255228e-05, + "loss": 0.0276, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.17435809969902039, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0281, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.1504661589860916, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0259, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.1678706258535385, + "learning_rate": 2.8221699881734e-05, + "loss": 0.029, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.13508610427379608, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0261, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12610375881195068, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0237, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.12272591143846512, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0231, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.1211056336760521, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0252, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.13204888999462128, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0276, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.11982017755508423, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0245, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.17140068113803864, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0236, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.12660211324691772, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0264, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1636001467704773, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0266, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.16939589381217957, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0276, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.15193505585193634, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0291, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16562964022159576, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0283, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.13603557646274567, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.024, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.14359496533870697, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0257, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.14077764749526978, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0247, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.14604462683200836, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0277, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.1240646094083786, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0262, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13144497573375702, + "learning_rate": 2.818639656220465e-05, + "loss": 0.0255, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.1384861171245575, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0227, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.16475403308868408, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0269, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.14310070872306824, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0258, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1498526632785797, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0263, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.13084106147289276, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0268, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.12004188448190689, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0252, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.14093606173992157, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0262, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.11016195267438889, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.0249, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.11101678013801575, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0246, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.1335095763206482, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0274, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1479913741350174, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0237, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.11868918687105179, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0231, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12799321115016937, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.027, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.17011503875255585, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0271, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.15915420651435852, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0274, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.11501206457614899, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0254, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.1457805633544922, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0274, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.1501278430223465, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0264, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.13190113008022308, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0272, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.14392758905887604, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0235, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.1375834047794342, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0241, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.14287824928760529, + "learning_rate": 2.814280752477756e-05, + "loss": 0.0269, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12234462797641754, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0259, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.1336476355791092, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0242, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.11315035820007324, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0245, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.14132073521614075, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0293, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.12053677439689636, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.0266, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.10896297544240952, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0268, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.11100424081087112, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.026, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.11256317049264908, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0237, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.14322060346603394, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0255, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.1655588448047638, + "learning_rate": 2.812283453196085e-05, + "loss": 0.026, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.14714404940605164, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0261, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.12317050993442535, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0247, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12872155010700226, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0258, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.1666097790002823, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0261, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.13425199687480927, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.16131597757339478, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0295, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.13762447237968445, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0242, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.12883082032203674, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0274, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.15029405057430267, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0269, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.13236507773399353, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0303, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.12054237723350525, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0218, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.1471109539270401, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0242, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.11926044523715973, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0275, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.13649292290210724, + "learning_rate": 2.809470494867301e-05, + "loss": 0.0231, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.15434928238391876, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0283, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.11536315083503723, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0271, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.12312797456979752, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0249, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.12750110030174255, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0235, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.1568569540977478, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0277, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.11667264997959137, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0258, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.14598897099494934, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0266, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.21723008155822754, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0258, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.14233075082302094, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0244, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.14836373925209045, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0265, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.146131694316864, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0258, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.16575932502746582, + "learning_rate": 2.807043872632495e-05, + "loss": 0.025, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.15142853558063507, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0245, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.11762700974941254, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0253, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.09714391082525253, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0233, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.1586424708366394, + "learning_rate": 2.806231820878714e-05, + "loss": 0.023, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.14317330718040466, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.0271, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.10633813589811325, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.0244, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.19832657277584076, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0249, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.14428019523620605, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0283, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.18295790255069733, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.0257, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.16204482316970825, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0285, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.14742101728916168, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0239, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1435142159461975, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0253, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.2000540792942047, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.16311310231685638, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0253, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.1238052025437355, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0249, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.16693037748336792, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0269, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.12382039427757263, + "learning_rate": 2.803581696238626e-05, + "loss": 0.025, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.13866499066352844, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0259, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.12111294269561768, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0258, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.13717366755008698, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0266, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.1119573786854744, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0264, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.1390918791294098, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.12881262600421906, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0251, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.1354016363620758, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0256, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.1285468488931656, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0315, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.17634150385856628, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0254, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.15077067911624908, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.0268, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16464342176914215, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0265, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.3138365149497986, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0255, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.15954458713531494, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0272, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15878041088581085, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0267, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.17743942141532898, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0273, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.16661205887794495, + "learning_rate": 2.800297040373295e-05, + "loss": 0.0268, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.13536596298217773, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.0241, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.13258172571659088, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0266, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.1524200588464737, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0276, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.12779775261878967, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0245, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1008654311299324, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0222, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.14413924515247345, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0297, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.14558148384094238, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0252, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.15591812133789062, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0248, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.12246204912662506, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0276, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.13106316328048706, + "learning_rate": 2.7982312890369e-05, + "loss": 0.0265, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.12663032114505768, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0234, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.1167084351181984, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15868185460567474, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0251, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13669976592063904, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0291, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.15100552141666412, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0237, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14283764362335205, + "learning_rate": 2.796987104810136e-05, + "loss": 0.024, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.13055892288684845, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0255, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.16733981668949127, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0244, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.1716628074645996, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0257, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.11838851124048233, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0288, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.14557431638240814, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0238, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.14981132745742798, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0254, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.11149631440639496, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0256, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.20500904321670532, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0255, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.12128293514251709, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0254, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.12934274971485138, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0254, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.1355874240398407, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0267, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13485173881053925, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0259, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.1481594294309616, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0251, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.15145331621170044, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0243, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.16746139526367188, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0284, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.10760390013456345, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0268, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.1460299789905548, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0264, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.15936219692230225, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0232, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.1373136043548584, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.0255, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.14508306980133057, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0243, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.153093621134758, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0262, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.15373852849006653, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0238, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.14033931493759155, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0267, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.1424216777086258, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0252, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.11820772290229797, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0253, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.13179944455623627, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0243, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.18073157966136932, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0256, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.12908786535263062, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0254, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.18455258011817932, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0263, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.14844198524951935, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0248, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.14291873574256897, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0268, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.105900838971138, + "learning_rate": 2.79029165248073e-05, + "loss": 0.027, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.14083242416381836, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0244, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.14579042792320251, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0239, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.13705924153327942, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0254, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.17981873452663422, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0225, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.13812783360481262, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0278, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12511365115642548, + "learning_rate": 2.789025068651819e-05, + "loss": 0.027, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.13009224832057953, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0245, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.145152747631073, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0281, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.1607820838689804, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0254, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.09400778263807297, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0258, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.1608203500509262, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0263, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.14968201518058777, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0247, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.1333402544260025, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0272, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.17871612310409546, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0256, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.10688412934541702, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0264, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.154439315199852, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0259, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.1270124316215515, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0279, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.15214042365550995, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0259, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.14094167947769165, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0261, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.11419129371643066, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.025, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.1447947919368744, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0258, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.12327348440885544, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0246, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.11033894866704941, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0259, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.128106951713562, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0247, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.13206088542938232, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0254, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.13664662837982178, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0274, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1349778026342392, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0262, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.10689893364906311, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0262, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.16241611540317535, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0269, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14955222606658936, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0251, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.13519948720932007, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0231, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.15076202154159546, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0272, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.15367662906646729, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0255, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.15819156169891357, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0242, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.12844862043857574, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0259, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.13392682373523712, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0245, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.15134917199611664, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.0287, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.14139734208583832, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0254, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.1562367081642151, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0246, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.1298898160457611, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0252, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13048380613327026, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0252, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.12535928189754486, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0271, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.19301816821098328, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0245, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.12262707948684692, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0262, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1394580751657486, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0272, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.21376213431358337, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0243, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.15442071855068207, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0285, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.13063938915729523, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0277, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.15023048222064972, + "learning_rate": 2.779844963973707e-05, + "loss": 0.025, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.17523249983787537, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0239, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.1312434822320938, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0234, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13352124392986298, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0249, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.12799884378910065, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0257, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.12704412639141083, + "learning_rate": 2.778765821684581e-05, + "loss": 0.025, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.11119242757558823, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0248, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.1564178764820099, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0282, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.13500694930553436, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0279, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.1542324423789978, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0258, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13888850808143616, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0281, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.14011257886886597, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.14887000620365143, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0254, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.15670792758464813, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0289, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.1288663148880005, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0242, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.13191169500350952, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.027, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.1320677250623703, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0239, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.13418279588222504, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0234, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.11019188910722733, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0232, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.13963419198989868, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0235, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.1163327619433403, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.028, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.13508521020412445, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0275, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.11098913848400116, + "learning_rate": 2.775078589916535e-05, + "loss": 0.025, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.16348809003829956, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0259, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.12986604869365692, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0246, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1302310824394226, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0259, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.11797303706407547, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.0252, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.1427793651819229, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0233, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.13411074876785278, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.0257, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.1528121531009674, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.11762527376413345, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0238, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.13639862835407257, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0257, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.1356336623430252, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0269, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.16879825294017792, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.13352762162685394, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0236, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.20271515846252441, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0296, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.11791368573904037, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0241, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14179861545562744, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.0256, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12654612958431244, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0243, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12589605152606964, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0222, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.12714968621730804, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0278, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.13410630822181702, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0248, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.12528881430625916, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0258, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.11806205660104752, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0308, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.14317870140075684, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0244, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14099767804145813, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0228, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.15186408162117004, + "learning_rate": 2.769825463131276e-05, + "loss": 0.0252, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.18532416224479675, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0262, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.1445159614086151, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0265, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14576950669288635, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0253, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1777651309967041, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0229, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.16440069675445557, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0234, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.12858232855796814, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0259, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.14329349994659424, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0254, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.1265982836484909, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0235, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13783425092697144, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0228, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.1327633559703827, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0248, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14867408573627472, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0257, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.1556752771139145, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0239, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.16407394409179688, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0261, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.14142248034477234, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0265, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.11142456531524658, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0221, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.127807155251503, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0265, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.14716863632202148, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0241, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.1455392986536026, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.024, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1665087193250656, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0253, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.16937066614627838, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0292, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.13188236951828003, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0254, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.1010027602314949, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.0236, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.15787169337272644, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.0279, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.169437438249588, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0246, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.22705122828483582, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0227, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.17403557896614075, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.027, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.1287415474653244, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0253, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.13813741505146027, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0238, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.11655540764331818, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.13267700374126434, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.025, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.10447681695222855, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0239, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.11865647882223129, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.025, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.13245069980621338, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0256, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.11377252638339996, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0266, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.11403302848339081, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0252, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.1518944650888443, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0286, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.13505733013153076, + "learning_rate": 2.761618148370765e-05, + "loss": 0.0253, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.12230793386697769, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0247, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.1600245237350464, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0267, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.13100600242614746, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.11071992665529251, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0254, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.16208025813102722, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.0287, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.1652526706457138, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.027, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14042513072490692, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.1674395203590393, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0257, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.138246089220047, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0272, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13758982717990875, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0227, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.15503156185150146, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0259, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.15050673484802246, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0279, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.13952329754829407, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.024, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.14903564751148224, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0242, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.09984514862298965, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.09952808916568756, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0233, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.14514309167861938, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0255, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.12824438512325287, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0232, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.15507018566131592, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0251, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.12279994785785675, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0225, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.13601353764533997, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0249, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.20022229850292206, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.0263, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.1407146006822586, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0226, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.1350381076335907, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0248, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.13220156729221344, + "learning_rate": 2.755998359318217e-05, + "loss": 0.0251, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.11746657639741898, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0256, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.10933379083871841, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0235, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.1263892501592636, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0244, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.14062181115150452, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.0255, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.1430615335702896, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0254, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.19069550931453705, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0254, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.16568611562252045, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0252, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.15911205112934113, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.0265, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.17054520547389984, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0265, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.15547682344913483, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0237, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.12778308987617493, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0242, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.1590787172317505, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0232, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13945519924163818, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.023, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.14507925510406494, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0247, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14541085064411163, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0245, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.17666012048721313, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0222, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.16692468523979187, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0218, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.1295796036720276, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0242, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.12527599930763245, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0273, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.13334719836711884, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0253, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.1175302267074585, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0251, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.14461147785186768, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0261, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.15658612549304962, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0236, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.15802229940891266, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0236, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.14738211035728455, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0252, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.1639637053012848, + "learning_rate": 2.750090542643057e-05, + "loss": 0.027, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1573086827993393, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0264, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.13114139437675476, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0238, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.1112227514386177, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.027, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.12209689617156982, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.13741743564605713, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0239, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14444704353809357, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0263, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.13822096586227417, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0246, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.14305464923381805, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0237, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13539855182170868, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0264, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.13791607320308685, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0234, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.205885648727417, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0234, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.1034194752573967, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0214, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.15530072152614594, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0258, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.1535874903202057, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0226, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.19778263568878174, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0238, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.175798699259758, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0255, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.1803887039422989, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0238, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.1906212419271469, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0253, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.13742655515670776, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.026, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.12772122025489807, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0261, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13805702328681946, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0259, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.14081542193889618, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0241, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.10023088753223419, + "learning_rate": 2.744810891043728e-05, + "loss": 0.0246, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.10111884027719498, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0215, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.11498598754405975, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0217, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10912947356700897, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.0257, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.13553178310394287, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0266, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.18210133910179138, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0234, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.11901862919330597, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0248, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.14622537791728973, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0234, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.1641179323196411, + "learning_rate": 2.742962757296782e-05, + "loss": 0.026, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.14369463920593262, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.0236, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.14231407642364502, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0213, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.10929716378450394, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0216, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.13811182975769043, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0241, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.13219693303108215, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0219, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.13322590291500092, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0259, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.14716853201389313, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0255, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.11750922352075577, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0272, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.1430707424879074, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1722290813922882, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.025, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.15502606332302094, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.0231, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.09124620258808136, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0246, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.16296057403087616, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0277, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.10881830751895905, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0222, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.11321897804737091, + "learning_rate": 2.739481230505444e-05, + "loss": 0.026, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.1362120658159256, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.03, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.14329490065574646, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0252, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.4512016475200653, + "learning_rate": 2.738782381719405e-05, + "loss": 0.0238, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12559202313423157, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0263, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.12616541981697083, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0248, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.1516106277704239, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0234, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.12785211205482483, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0264, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16447393596172333, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.026, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.1648014783859253, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0264, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.14414788782596588, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0249, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.10982059687376022, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0254, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.12677313387393951, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0263, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.12355965375900269, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0223, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.15190085768699646, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0246, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.13910633325576782, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0231, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.11962535977363586, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0259, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.1566285789012909, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0256, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.12282316386699677, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0253, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.16674686968326569, + "learning_rate": 2.735040900116229e-05, + "loss": 0.025, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.14292193949222565, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0277, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.1504344791173935, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0268, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.16317704319953918, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0229, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.20625433325767517, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0242, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.1895458698272705, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.024, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.1987384408712387, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.0258, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.18668580055236816, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.1807073950767517, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0246, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.11927719414234161, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.028, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.11955022811889648, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0259, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.15261328220367432, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0268, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.15620580315589905, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0254, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.12000948935747147, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0234, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.1188279539346695, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.0248, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.13279424607753754, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0241, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.13379216194152832, + "learning_rate": 2.731275407491117e-05, + "loss": 0.026, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.12054166197776794, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0215, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.15941590070724487, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0279, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.18678002059459686, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0263, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.21807490289211273, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.0239, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.12243734300136566, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.028, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.14628900587558746, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0248, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.17471152544021606, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0257, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.11189062893390656, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0242, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.16700562834739685, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0252, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.1213584691286087, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0251, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.14496786892414093, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0258, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.17899028956890106, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0248, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13594108819961548, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0246, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.1365566849708557, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.021, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.10815171152353287, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0227, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.11390877515077591, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0209, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.1541927456855774, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0213, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.1342182159423828, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0251, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.1300208568572998, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0249, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14917100965976715, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0274, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.14349427819252014, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.0269, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10514543205499649, + "learning_rate": 2.72605878418951e-05, + "loss": 0.024, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.12946246564388275, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0234, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.1120578870177269, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.0246, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.14252731204032898, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.0287, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.0984109416604042, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0234, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.13058412075042725, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0235, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.1468561738729477, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0263, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.1297929584980011, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0224, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.15071524679660797, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0254, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.15211531519889832, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0239, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.13758648931980133, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0254, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.13972750306129456, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0239, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1762881577014923, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0235, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.13277511298656464, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.025, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.11352386325597763, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0225, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.1340995877981186, + "learning_rate": 2.722476145603374e-05, + "loss": 0.0232, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.16372360289096832, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0224, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.130090594291687, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0231, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.10481512546539307, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0229, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.1453929841518402, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0255, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.133439838886261, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0248, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.1380738615989685, + "learning_rate": 2.721037237880308e-05, + "loss": 0.023, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.1386265754699707, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0264, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.14298051595687866, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.0247, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.11756420135498047, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0222, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1197751984000206, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0216, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.2194000482559204, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0243, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.14157775044441223, + "learning_rate": 2.719594991888059e-05, + "loss": 0.023, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.14389759302139282, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0228, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.12442908436059952, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0267, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.1188371554017067, + "learning_rate": 2.718872618272836e-05, + "loss": 0.027, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11137005686759949, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0237, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.10907725989818573, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0251, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.150213360786438, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0229, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.2098330855369568, + "learning_rate": 2.717908157623383e-05, + "loss": 0.0245, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.12558938562870026, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0257, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.13659170269966125, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0247, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.12037219852209091, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0228, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.13085514307022095, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0252, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.13916544616222382, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0251, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.13098886609077454, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.021, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.12777045369148254, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0258, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.10204357653856277, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0249, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.11189272999763489, + "learning_rate": 2.715732712163776e-05, + "loss": 0.0232, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.1412421613931656, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0255, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.09470678865909576, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0217, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.09736761450767517, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0233, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12988215684890747, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.1080017015337944, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0263, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.12020932137966156, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0238, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.1058862954378128, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0247, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11686211824417114, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.024, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.11078432947397232, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0226, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.13394472002983093, + "learning_rate": 2.713306780076029e-05, + "loss": 0.0237, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.11299733072519302, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.025, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.1196199283003807, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0251, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.16477037966251373, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0241, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.1175335943698883, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.11024156212806702, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0237, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.12922869622707367, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0278, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.1429518610239029, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0254, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.14210145175457, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0225, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.12270870059728622, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0257, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.12436466664075851, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0258, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.10704488307237625, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0224, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.15890415012836456, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0252, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.12372975796461105, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0225, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.11652565747499466, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.028, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.1483478546142578, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0255, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.13188686966896057, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0238, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.18725869059562683, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0244, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.15949182212352753, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0225, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.14752954244613647, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0229, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.14210951328277588, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0233, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.1388787031173706, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.0253, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.11010867357254028, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0227, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.13336122035980225, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0227, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.10118173807859421, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.0265, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.1434611827135086, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0263, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.14133326709270477, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.026, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.15359432995319366, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0236, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.12927602231502533, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.022, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.10330940783023834, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0236, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.1264553815126419, + "learning_rate": 2.705973772304842e-05, + "loss": 0.025, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.15007789433002472, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0267, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.11090201884508133, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0257, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.12769529223442078, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0252, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.0977141410112381, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0244, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.14598220586776733, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0254, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.11231301724910736, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0235, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.09610529243946075, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0262, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.1549275815486908, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0234, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.11284032464027405, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.0217, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.14465086162090302, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.0235, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.13792403042316437, + "learning_rate": 2.703264323632037e-05, + "loss": 0.024, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.1438492238521576, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0239, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.12613703310489655, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0244, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.14072397351264954, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0225, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.1420842856168747, + "learning_rate": 2.702276326859174e-05, + "loss": 0.026, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.17447218298912048, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0243, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.1708984673023224, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0259, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.1555069535970688, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.0252, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.12473946064710617, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0227, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.15750949084758759, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0248, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.14009647071361542, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0267, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.1416911780834198, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0229, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.11944608390331268, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0235, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.126204252243042, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0268, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.1186814159154892, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0242, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.12699876725673676, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0239, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.10452742129564285, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0221, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.1422647386789322, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0247, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.11601094901561737, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0232, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12339281290769577, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0223, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1883407086133957, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0243, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.1730756312608719, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0294, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.10429747402667999, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0194, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.1358909010887146, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0217, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.14417894184589386, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0265, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.12299799919128418, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0236, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.1361788511276245, + "learning_rate": 2.696816263672279e-05, + "loss": 0.024, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.14492252469062805, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0241, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.1209225058555603, + "learning_rate": 2.696317709902802e-05, + "loss": 0.0239, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.11048539727926254, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.12423776835203171, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0243, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.1254356950521469, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.12912537157535553, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0254, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.10864657163619995, + "learning_rate": 2.695069735893089e-05, + "loss": 0.022, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.15519478917121887, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0253, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.15276901423931122, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0231, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.1401076465845108, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0271, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.14237190783023834, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.024, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1186559721827507, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0238, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.1254940629005432, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0233, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.1337885558605194, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0255, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12958411872386932, + "learning_rate": 2.69306825906724e-05, + "loss": 0.025, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.11610005795955658, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0238, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.13515442609786987, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0239, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1492868959903717, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0224, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.1293300986289978, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.021, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12354710698127747, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.13226614892482758, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.0247, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.1368253529071808, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0255, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.13044875860214233, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0228, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14107753336429596, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0221, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.13521252572536469, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0236, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.13718804717063904, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0217, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.15192179381847382, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.0247, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.13718710839748383, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0244, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11781664192676544, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0239, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.14431379735469818, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0248, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.15059512853622437, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0231, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.12076479196548462, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0242, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.19006308913230896, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0238, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.140592560172081, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0233, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.1528412401676178, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0229, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.16418038308620453, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0242, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.12185576558113098, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0254, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11594245582818985, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0248, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1216709092259407, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0234, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.12024594098329544, + "learning_rate": 2.686776313743295e-05, + "loss": 0.024, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.1321716010570526, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.0255, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.13785192370414734, + "learning_rate": 2.686270520700719e-05, + "loss": 0.025, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.12321551889181137, + "learning_rate": 2.6860174890357e-05, + "loss": 0.0233, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14606279134750366, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0249, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.17512187361717224, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0272, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.14974310994148254, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.0243, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.11758870631456375, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0241, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.11157143115997314, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0232, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.10899091511964798, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0231, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.1136092096567154, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0224, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.10306857526302338, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0211, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13241226971149445, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0222, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.11886361986398697, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0244, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.11962216347455978, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0237, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.14516247808933258, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.14592237770557404, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0253, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1278424859046936, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0233, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.12451537698507309, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0218, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.11746378242969513, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0266, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.14618739485740662, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0226, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.1512313336133957, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0233, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.14035750925540924, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0264, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.13232699036598206, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.023, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.14616237580776215, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.0233, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.12882372736930847, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0265, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.13778620958328247, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.024, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.15017054975032806, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0239, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.14781621098518372, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0215, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1670294851064682, + "learning_rate": 2.679407107451536e-05, + "loss": 0.022, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.1112075075507164, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0232, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.12219138443470001, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0202, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12922540307044983, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.11967732757329941, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0226, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.1377563774585724, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0224, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.14771930873394012, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0234, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.14800170063972473, + "learning_rate": 2.677617030053916e-05, + "loss": 0.0251, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.1233784481883049, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0232, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.16905057430267334, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0214, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.13751909136772156, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0248, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.13289029896259308, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0216, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.10788674652576447, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0216, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.13367260992527008, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1270209699869156, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0239, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.140412375330925, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0218, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.17788565158843994, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0253, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.13695330917835236, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.023, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.1250872015953064, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0232, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.11347977817058563, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0243, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.15546919405460358, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0233, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.13752058148384094, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.10431966185569763, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0215, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.17878347635269165, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0262, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.13218404352664948, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0241, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.12244930118322372, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.0247, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.11582386493682861, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0219, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.1160157322883606, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0203, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.10731764137744904, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0234, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.11152671277523041, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0218, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.12760375440120697, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0218, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.13706916570663452, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0219, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.14939993619918823, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0208, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.1169784739613533, + "learning_rate": 2.67092985767871e-05, + "loss": 0.0217, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.1569128930568695, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12233320623636246, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0251, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13152818381786346, + "learning_rate": 2.670154386126402e-05, + "loss": 0.0231, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.1529632955789566, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0246, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.11001774668693542, + "learning_rate": 2.669636960731296e-05, + "loss": 0.0223, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.1473267376422882, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.025, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.12853027880191803, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0246, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.15698745846748352, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0209, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12175270169973373, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0229, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.15634675323963165, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.0224, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.16060872375965118, + "learning_rate": 2.668082553346281e-05, + "loss": 0.0254, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.13773131370544434, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0224, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.14258436858654022, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0216, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.13032476603984833, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0215, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.17300327122211456, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0221, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.13187018036842346, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0226, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.14065369963645935, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0232, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.11484711617231369, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0263, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1380661576986313, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0209, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.17800134420394897, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0244, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.14836716651916504, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0277, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14563506841659546, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.023, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.11463280767202377, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.026, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.13092686235904694, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0222, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.12265385687351227, + "learning_rate": 2.664443191002912e-05, + "loss": 0.025, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.14121955633163452, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0242, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.12938837707042694, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0222, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.1922816038131714, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.0228, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.12596914172172546, + "learning_rate": 2.663400187275054e-05, + "loss": 0.022, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.14095240831375122, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0244, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.1080891340970993, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0228, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.13916714489459991, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.0239, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.1489475965499878, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0251, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.11781706660985947, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0207, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.11544635146856308, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0232, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.14827974140644073, + "learning_rate": 2.661571529941175e-05, + "loss": 0.0236, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.12821446359157562, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.0186, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.1303764134645462, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0234, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.1049366369843483, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0232, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.15925322473049164, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0275, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.13151033222675323, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0213, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.09425143152475357, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0215, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.15108847618103027, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0228, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.1374080330133438, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0231, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.16061817109584808, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0253, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.1707434207201004, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0259, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.12051897495985031, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0217, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.09191743284463882, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0224, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.12547235190868378, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0245, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14491257071495056, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0228, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.11896242201328278, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0239, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.14253611862659454, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0233, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.12803201377391815, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.0226, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.12893305718898773, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0231, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.16127154231071472, + "learning_rate": 2.656586262343925e-05, + "loss": 0.0234, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.13307608664035797, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0258, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.11211924254894257, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.0219, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.1672789305448532, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0248, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.12878142297267914, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0231, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.11429315060377121, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0233, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.13469499349594116, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0254, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.12753085792064667, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0258, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.11875704675912857, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0234, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.13718342781066895, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0223, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.10762278735637665, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.021, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.11651597917079926, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0222, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.12953802943229675, + "learning_rate": 2.653421326308085e-05, + "loss": 0.0235, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.14820824563503265, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.023, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.16258545219898224, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0266, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.14097236096858978, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0224, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.1039188802242279, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0216, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15014079213142395, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0246, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12850438058376312, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.023, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.12245986610651016, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0209, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.15426407754421234, + "learning_rate": 2.651304359235998e-05, + "loss": 0.025, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.13411583006381989, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0229, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.13705122470855713, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0254, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.13120216131210327, + "learning_rate": 2.650509053370553e-05, + "loss": 0.022, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14599482715129852, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0249, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.11628340184688568, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0215, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.13643144071102142, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0228, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.12846967577934265, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0214, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.13181492686271667, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0245, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1277213841676712, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0244, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12288910150527954, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0226, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.14789454638957977, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0215, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.11457780748605728, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0248, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.12511487305164337, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0206, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.12255935370922089, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0245, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.13214053213596344, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0238, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.1601077914237976, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.027, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.1244364008307457, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0212, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.11869437992572784, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0239, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.14327670633792877, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0227, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.12729141116142273, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0248, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.10953658074140549, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0205, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.20326738059520721, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0225, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.15614672005176544, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0237, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.11863692849874496, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0208, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.10683806240558624, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0224, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.1529647260904312, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.024, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.18428578972816467, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0232, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.19935141503810883, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.0223, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.14675194025039673, + "learning_rate": 2.643583521046136e-05, + "loss": 0.0216, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.21206484735012054, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0243, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.14521193504333496, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0214, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.11565426737070084, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0236, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.1087750792503357, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0215, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.11942578107118607, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0229, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.138235405087471, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0219, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.1079302579164505, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0222, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.15928085148334503, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0225, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.10920153558254242, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.024, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.12247785925865173, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0205, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.10454121232032776, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0238, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.1726839691400528, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0211, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1617811918258667, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0225, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.15452679991722107, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0214, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1573309749364853, + "learning_rate": 2.639561293166059e-05, + "loss": 0.022, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.13596323132514954, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0234, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.09188699722290039, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0219, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.12158972024917603, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0202, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.16615170240402222, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.0221, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.1163380891084671, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0255, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.17704762518405914, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0243, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.1591942012310028, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.021, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.14697043597698212, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0241, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1234351247549057, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0261, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.1379440575838089, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0266, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1025586649775505, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.022, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.11894206702709198, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0226, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1431341916322708, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0241, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.10291697829961777, + "learning_rate": 2.635789644394928e-05, + "loss": 0.022, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.1622016727924347, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0229, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.11867184937000275, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0262, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.18177814781665802, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0237, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.15159954130649567, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0218, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.11481288820505142, + "learning_rate": 2.634438526911474e-05, + "loss": 0.024, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.10989677160978317, + "learning_rate": 2.634168044873305e-05, + "loss": 0.024, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.12370797991752625, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0235, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1390318125486374, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0231, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.14030151069164276, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.0207, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11659140884876251, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.021, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.16727864742279053, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0242, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.11524021625518799, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0265, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.10776443779468536, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0224, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.16366012394428253, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0264, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.1102370023727417, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0245, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11543556302785873, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0216, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.11101235449314117, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0208, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.1257167011499405, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.024, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.1063276082277298, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0241, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.14303457736968994, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0242, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.12862786650657654, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0244, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15487778186798096, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.14244592189788818, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0211, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1066693440079689, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0199, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11721482872962952, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0214, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.13803230226039886, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.0227, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.11225558817386627, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.028, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.11660541594028473, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0239, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12551093101501465, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0221, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.1305139809846878, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.1281510442495346, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0235, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.14735114574432373, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0213, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.10699455440044403, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0231, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.16579580307006836, + "learning_rate": 2.626559653214893e-05, + "loss": 0.0213, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.09389153122901917, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0218, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.17565062642097473, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0245, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.1368194818496704, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0214, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.10599412024021149, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0221, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.12645669281482697, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0233, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.12325828522443771, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.0228, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.11353788524866104, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.10889376699924469, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0227, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.12935040891170502, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0214, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.1422247439622879, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0219, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.16723784804344177, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0239, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.09775881469249725, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.0229, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.13455381989479065, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0253, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.13700921833515167, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0239, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12252738326787949, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0231, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.15983854234218597, + "learning_rate": 2.622181856488204e-05, + "loss": 0.0223, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.18716242909431458, + "learning_rate": 2.621907518918971e-05, + "loss": 0.025, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.16380536556243896, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0229, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.1524866819381714, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0221, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.12281060218811035, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0213, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.09793791174888611, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0249, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.12271476536989212, + "learning_rate": 2.620534553467634e-05, + "loss": 0.0221, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.12616175413131714, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0236, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.10274055600166321, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0205, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.16873067617416382, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.027, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.10157257318496704, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0236, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.11863094568252563, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0225, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.09981630742549896, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0233, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11045375466346741, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0226, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.13111108541488647, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0232, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.12417030334472656, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0226, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.12029916048049927, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0217, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.10766672343015671, + "learning_rate": 2.617506544598029e-05, + "loss": 0.025, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.12039721757173538, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0228, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.11044551432132721, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0223, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.12463348358869553, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.0232, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.14382848143577576, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0238, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.13677646219730377, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0239, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.11536712944507599, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0224, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.12461430579423904, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0253, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.16901540756225586, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0242, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.12255802005529404, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0212, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.14185196161270142, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.0218, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.1292722374200821, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0249, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12001507729291916, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0222, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.12286248803138733, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0211, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.12487491965293884, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0231, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13417202234268188, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0257, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.11690660566091537, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0219, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15474916994571686, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0223, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.1720103919506073, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0239, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.16786889731884003, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0216, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.2279045283794403, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.0254, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13872499763965607, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0238, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.1881643831729889, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0211, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.1372661143541336, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0237, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.10514581203460693, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0217, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.11103228479623795, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0248, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1349911093711853, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0237, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.12774604558944702, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0245, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.15125125646591187, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0233, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.1190832257270813, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0248, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.1088283434510231, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0187, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.12167634069919586, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.023, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.15135347843170166, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0226, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.12857669591903687, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.0195, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.13588359951972961, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0234, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.14902308583259583, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0235, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.15095068514347076, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0209, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.1116727739572525, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0255, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11652971059083939, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0211, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.11875079572200775, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0223, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.11100098490715027, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.022, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.13648055493831635, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0234, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.12870723009109497, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0205, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.10456091910600662, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0232, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.13770641386508942, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0212, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.1526155173778534, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0244, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12664297223091125, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0241, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.12000498175621033, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0239, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.11930232495069504, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.0207, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.11384909600019455, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0217, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.1746685951948166, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.023, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.15868088603019714, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0233, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.15131928026676178, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0194, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.10279352962970734, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0215, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.13810355961322784, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0202, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.13024795055389404, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0217, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1374605894088745, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0215, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.14356477558612823, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0206, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.11046172678470612, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0236, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.10146678239107132, + "learning_rate": 2.601090850395125e-05, + "loss": 0.021, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13247323036193848, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0217, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.13672757148742676, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0201, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.08651759475469589, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.0196, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.14821568131446838, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0223, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11985234171152115, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0208, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11231720447540283, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0219, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.12338588386774063, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0202, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.15079368650913239, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0229, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.1348753422498703, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0209, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1333470344543457, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0207, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.10075798630714417, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0232, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1217522993683815, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0247, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.12041056901216507, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.022, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.10997799783945084, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0226, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.119750015437603, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0239, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.15246307849884033, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0239, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.1431925892829895, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0227, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.1181643158197403, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0227, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.13645906746387482, + "learning_rate": 2.595742267991401e-05, + "loss": 0.0213, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.12661530077457428, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0233, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.1344967484474182, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.0221, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.13433857262134552, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0238, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.10517163574695587, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0237, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.10689525306224823, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.022, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.13307900726795197, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0209, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.14696283638477325, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0261, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.12473653256893158, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0213, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.106630340218544, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0221, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.12092334777116776, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0232, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.11033511906862259, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0193, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11916322261095047, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0212, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.1204444020986557, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0228, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.1373489499092102, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0225, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.13124211132526398, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0208, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.18858365714550018, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0258, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.1558879315853119, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0228, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.1481737494468689, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0242, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.10424906760454178, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.021, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.15494640171527863, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0238, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.1801268309354782, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0238, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.14176428318023682, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0228, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.15850430727005005, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0235, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.14727924764156342, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0232, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.14907293021678925, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0242, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.11011895537376404, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0231, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.13541805744171143, + "learning_rate": 2.588090010443142e-05, + "loss": 0.0224, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.13078846037387848, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0228, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.14508767426013947, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0207, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.12445177882909775, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0227, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.14860141277313232, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0244, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.13311992585659027, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0207, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.1157015860080719, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0209, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.13310302793979645, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.0234, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.14026974141597748, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0246, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.14788632094860077, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0213, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.1551811844110489, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0241, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.10706475377082825, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.021, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.12536053359508514, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0228, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.1260751187801361, + "learning_rate": 2.584384079351628e-05, + "loss": 0.022, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.1082935482263565, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0207, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.13922622799873352, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0251, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1240227147936821, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0223, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.14512665569782257, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0218, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12859801948070526, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.0208, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.13418391346931458, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.021, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.10546918958425522, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.022, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.10668392479419708, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10407276451587677, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.02, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.13980497419834137, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0235, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.16233955323696136, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0224, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.16743852198123932, + "learning_rate": 2.580950866038058e-05, + "loss": 0.022, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.09152300655841827, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0206, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.14319349825382233, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0205, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.1045396700501442, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0219, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.10865644365549088, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0201, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.12166012078523636, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0217, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.12505589425563812, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0224, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.1353408247232437, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0242, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.1379590630531311, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0221, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12337373197078705, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0235, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.12092723697423935, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0233, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.09944337606430054, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0215, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.15161453187465668, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0193, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.0963459238409996, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0225, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.15534570813179016, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0247, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.142191082239151, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0228, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.1454230695962906, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0226, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.12661893665790558, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0211, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.11373333632946014, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0213, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12031935155391693, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.0208, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.13401862978935242, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0224, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.11285468935966492, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0231, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.12453637272119522, + "learning_rate": 2.574625966768285e-05, + "loss": 0.0211, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.13041555881500244, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1391122043132782, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0224, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.1345762014389038, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0244, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.12814867496490479, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0219, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.11058632284402847, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0191, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.15668845176696777, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.021, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.1509203314781189, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.0254, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.13209059834480286, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0201, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.1357070952653885, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0226, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.0896834284067154, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0207, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.1570383608341217, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0234, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.16916486620903015, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0254, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.1385425180196762, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.0218, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.11236828565597534, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0214, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11224918812513351, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0233, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.1305687576532364, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0219, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.15051157772541046, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0226, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.13418416678905487, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0244, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.14324583113193512, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0235, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.10810746997594833, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0197, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.14513953030109406, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0215, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1187531128525734, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0193, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11865831911563873, + "learning_rate": 2.567971342427169e-05, + "loss": 0.021, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.12040061503648758, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0229, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.10597001016139984, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0209, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.1219123974442482, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0242, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.11732123047113419, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.0203, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.15642790496349335, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0226, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.1893257051706314, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0247, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11829031258821487, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0239, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.14971555769443512, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0253, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.12424097210168839, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0251, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1071329414844513, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0271, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.11146610230207443, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0214, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.10925749689340591, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.0212, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.14797845482826233, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0221, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.10348086059093475, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0218, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.14607182145118713, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0217, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.11097954958677292, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0259, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.11300431936979294, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0197, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.11342521011829376, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0202, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.12889035046100616, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0217, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.10449834913015366, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0216, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1293669193983078, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0227, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.10667148977518082, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.021, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.0839434340596199, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0236, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.12906265258789062, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0223, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.1100582480430603, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0243, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.10644520074129105, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.0215, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.11693669110536575, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.022, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.13665053248405457, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0227, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12752027809619904, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0241, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.11534811556339264, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0234, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.1467696875333786, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0245, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10596702992916107, + "learning_rate": 2.558641420481684e-05, + "loss": 0.02, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.14685864746570587, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.022, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.13554386794567108, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.026, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.12572740018367767, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0232, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.1583555042743683, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0209, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.12913484871387482, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0226, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.11418721824884415, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0225, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.12722279131412506, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.0211, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.1382816582918167, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.0243, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.11612103134393692, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0212, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.1537163406610489, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.021, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.10462892800569534, + "learning_rate": 2.555415209239832e-05, + "loss": 0.0214, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.12759479880332947, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0232, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.1290607452392578, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0222, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12458004057407379, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0213, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.1113182008266449, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0223, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.15024979412555695, + "learning_rate": 2.553945541728904e-05, + "loss": 0.022, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.12657828629016876, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.025, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.13719463348388672, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.024, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.13465221226215363, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0254, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.08891251683235168, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0215, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.08946283161640167, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0224, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.13241387903690338, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0229, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.13068708777427673, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0215, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.12115749716758728, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.025, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.12863430380821228, + "learning_rate": 2.551295099554365e-05, + "loss": 0.0202, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.1332341581583023, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0204, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.102240651845932, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0199, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.16963273286819458, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0226, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12840960919857025, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0225, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.09958021342754364, + "learning_rate": 2.549819836273345e-05, + "loss": 0.02, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.10737278312444687, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0226, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.11473958194255829, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0219, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.12188851088285446, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0216, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.1229034960269928, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0227, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.1316317766904831, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1230575442314148, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0218, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13473588228225708, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0221, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1293632537126541, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0223, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.11663107573986053, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0198, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.147003173828125, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0216, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.10070289671421051, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0204, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12855689227581024, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0222, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.10310211777687073, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0227, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.25780248641967773, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0204, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.11621664464473724, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0196, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.15240156650543213, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0232, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.1359490305185318, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0191, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.13072653114795685, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.0242, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.1296282559633255, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0214, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.12183726578950882, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0215, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.10602327436208725, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0222, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.12424091249704361, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0215, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.14287705719470978, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0217, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.10017845034599304, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0214, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.1265304833650589, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0214, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.15279516577720642, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0194, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.11116814613342285, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0197, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.14380785822868347, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0187, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.12522602081298828, + "learning_rate": 2.541224092500406e-05, + "loss": 0.023, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.12386767566204071, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.0197, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11278095096349716, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0205, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12119927257299423, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0206, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.14128951728343964, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0197, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.12576138973236084, + "learning_rate": 2.539735331078111e-05, + "loss": 0.024, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.13276934623718262, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0217, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.15223100781440735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0222, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.13918957114219666, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0211, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.10129881650209427, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0192, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.15048135817050934, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0191, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.11579657346010208, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.019, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.1882474422454834, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0241, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.12608706951141357, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.021, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.14131402969360352, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0234, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.11546603590250015, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0208, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.0989922359585762, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0219, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.13154740631580353, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0234, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.12450987845659256, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0219, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.15696841478347778, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0228, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12059786915779114, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0236, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.1058221235871315, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0234, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.11429761350154877, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0232, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.11841994524002075, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0193, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.12387857586145401, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0239, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.11453676968812943, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0224, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.15571972727775574, + "learning_rate": 2.533461009735874e-05, + "loss": 0.022, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.10241609811782837, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.022, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.11682361364364624, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0211, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.13623486459255219, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0215, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.12484022229909897, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0212, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.09105533361434937, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0223, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.1024160087108612, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0174, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.14108790457248688, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0215, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.1268249899148941, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0211, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12255556881427765, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0192, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.15904900431632996, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0259, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.1356613039970398, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.023, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.12410393357276917, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0223, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1291753053665161, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0221, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.109084852039814, + "learning_rate": 2.529258887162501e-05, + "loss": 0.0243, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.16595998406410217, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0201, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.14062269032001495, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0232, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.13198593258857727, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0214, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.10668856650590897, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0229, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.11266572028398514, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0233, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.12742944061756134, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0221, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.12420892715454102, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0215, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.1265600174665451, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0215, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.11914686858654022, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0222, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.12876294553279877, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0233, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.09800750762224197, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0219, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1273864060640335, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0238, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.13299106061458588, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0226, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.09396592527627945, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.12955208122730255, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0204, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.16461864113807678, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0211, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.11535970121622086, + "learning_rate": 2.524135719213194e-05, + "loss": 0.02, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.11860322207212448, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0212, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10968036204576492, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0207, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.12494095414876938, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.02, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.10398012399673462, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0205, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.12939494848251343, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0216, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.10996326059103012, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0199, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.13704033195972443, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0208, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.11454308778047562, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.0212, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.14186057448387146, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0226, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.12624409794807434, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0199, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.14392296969890594, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.15774424374103546, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0213, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.1352480947971344, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0192, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.10594955086708069, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0231, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.11280091851949692, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0202, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11683343350887299, + "learning_rate": 2.519293379145164e-05, + "loss": 0.022, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.11721377819776535, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0223, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.18298691511154175, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.0225, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12927964329719543, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0202, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.1413726508617401, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.021, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.1330733746290207, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0228, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12341104447841644, + "learning_rate": 2.517472387591168e-05, + "loss": 0.021, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.12199880182743073, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.0202, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.12259527295827866, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.0227, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.11682213097810745, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0246, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.12783272564411163, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0204, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.09663784503936768, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0228, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.13686567544937134, + "learning_rate": 2.515648614306516e-05, + "loss": 0.02, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.12164216488599777, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0234, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.1371183544397354, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0217, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.12031334638595581, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0203, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.1183433011174202, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.1265946477651596, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0224, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.12686678767204285, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0202, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.13161398470401764, + "learning_rate": 2.513517369661726e-05, + "loss": 0.0216, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.13780610263347626, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0205, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.14213374257087708, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0237, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.12532930076122284, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0215, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1276065707206726, + "learning_rate": 2.512297821731217e-05, + "loss": 0.02, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.08893107622861862, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12819649279117584, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0202, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.09707193821668625, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0202, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.511077043763402e-05, + "loss": 0.0208, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.10047636926174164, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0219, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.15637417137622833, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0218, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.15387912094593048, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0213, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.14876650273799896, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0222, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.13631325960159302, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0214, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.12577500939369202, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0232, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15896311402320862, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0237, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.16459719836711884, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0223, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11081109195947647, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0223, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.12327340245246887, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0219, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14812195301055908, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0228, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.11862681806087494, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0252, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.10684223473072052, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0227, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.1279568374156952, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0216, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.13341103494167328, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0213, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.11913862824440002, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0214, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.13493330776691437, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.0214, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.1272144317626953, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0209, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.11348695307970047, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0217, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.16526950895786285, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0198, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.14953579008579254, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0231, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.12980960309505463, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0226, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.16652929782867432, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0244, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1722235530614853, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.026, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.23629626631736755, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0221, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.12417787313461304, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.023, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.11768155544996262, + "learning_rate": 2.502804732645032e-05, + "loss": 0.0186, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.13731959462165833, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.023, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.18034201860427856, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.0201, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11240512877702713, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0217, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.10463474690914154, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0201, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.1273874044418335, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0194, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.12314468622207642, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0237, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.13170552253723145, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.021, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.12725704908370972, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0198, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.11761748790740967, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0198, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13019445538520813, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0237, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.09089519083499908, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.0202, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.13669925928115845, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0236, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.0891583263874054, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0189, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.12015428394079208, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.0214, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.12889859080314636, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0214, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.1092396154999733, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.022, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.1011805608868599, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0201, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.11652060598134995, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.019, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.12280789762735367, + "learning_rate": 2.496950169240971e-05, + "loss": 0.0229, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.1253744661808014, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0245, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.15854160487651825, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0257, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.11671199649572372, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0204, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.15195102989673615, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.11878954619169235, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0207, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.1035882979631424, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0195, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12105780094861984, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0191, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.13278989493846893, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0207, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.16576215624809265, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.0228, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.13104894757270813, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0197, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.14476071298122406, + "learning_rate": 2.49354818242132e-05, + "loss": 0.023, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.14116041362285614, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0204, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.133507639169693, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0213, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11147522926330566, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0192, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.1317354440689087, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0229, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.15885432064533234, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0214, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.0996033102273941, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0213, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.11059842258691788, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0181, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.1103382259607315, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0252, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.12147006392478943, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.1167205348610878, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0217, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.11838674545288086, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0223, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.1745499074459076, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0213, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.09892687201499939, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0213, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.17135944962501526, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0228, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.13930797576904297, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0211, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.10764696449041367, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0204, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11211943626403809, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0207, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1359340399503708, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0215, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.11819283664226532, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0213, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.1166793555021286, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0216, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.14411170780658722, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0206, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.1380920261144638, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0211, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.12901239097118378, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.0224, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.1171313151717186, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.0207, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.15169396996498108, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0225, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.2066943198442459, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0182, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1179543286561966, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0226, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.12273375689983368, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.0224, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.09459412097930908, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0177, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.10208278894424438, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0207, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.11947600543498993, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.0246, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10635806620121002, + "learning_rate": 2.483599695058918e-05, + "loss": 0.021, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.15165211260318756, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0242, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.10573819279670715, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0221, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.10153646767139435, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0212, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.12596352398395538, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.021, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.11259406059980392, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0214, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.09709548950195312, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0186, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1462256908416748, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0228, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.15480898320674896, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0245, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.10707002878189087, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0197, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.10963082313537598, + "learning_rate": 2.480475069778109e-05, + "loss": 0.019, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.13645771145820618, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0229, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.08208916336297989, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0194, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.11223483830690384, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0212, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.12233787775039673, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0228, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.12175288051366806, + "learning_rate": 2.478909963381241e-05, + "loss": 0.0204, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.14267505705356598, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0203, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.12649142742156982, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0217, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.11254285275936127, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0214, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.1454164832830429, + "learning_rate": 2.477656539969673e-05, + "loss": 0.0221, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.17177407443523407, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0212, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.1089746281504631, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0222, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.1525813490152359, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0236, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.09651731699705124, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0214, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.12710145115852356, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0233, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.15999987721443176, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0215, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1600969433784485, + "learning_rate": 2.475460191424065e-05, + "loss": 0.0201, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.11045356094837189, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0208, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.14192573726177216, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.0216, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.11450321227312088, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0202, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.13454747200012207, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.021, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.11518846452236176, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0194, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.10673099011182785, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0214, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.12131883949041367, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0203, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.18098120391368866, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0182, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.09981405735015869, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0214, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.09194957464933395, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0209, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.11133445054292679, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0227, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.14570456743240356, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0181, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.14878204464912415, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0213, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.1475936472415924, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0229, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.14091694355010986, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0229, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.10411489754915237, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.021, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.10015816241502762, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0211, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09339930862188339, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0206, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.13016946613788605, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0223, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.10682770609855652, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0258, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.13658446073532104, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0232, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.12742926180362701, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.13928014039993286, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0208, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.10824346542358398, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0199, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.11346063017845154, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.0225, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12237603217363358, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0226, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.12201637774705887, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0209, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.09991955012083054, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0239, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.15387621521949768, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.025, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.1588744819164276, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0192, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.1347019523382187, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0196, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.12746664881706238, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0212, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11522521078586578, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0197, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.09697534143924713, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0219, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12183485180139542, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0195, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.13119301199913025, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0195, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.10748504102230072, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0192, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.1617487072944641, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0224, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.11518653482198715, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0207, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.17646238207817078, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0233, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.13729365170001984, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.09777814894914627, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0207, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.117672860622406, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0254, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.09014946967363358, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0214, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.1273731142282486, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0223, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10932891070842743, + "learning_rate": 2.460937119575347e-05, + "loss": 0.0224, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.1672651767730713, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0205, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.0930962786078453, + "learning_rate": 2.46030216185615e-05, + "loss": 0.02, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.09708375483751297, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0192, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.13240690529346466, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0214, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.10310725122690201, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0204, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.13352994620800018, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0206, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.15277482569217682, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0222, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.14223086833953857, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0229, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12845414876937866, + "learning_rate": 2.458077513854331e-05, + "loss": 0.0221, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13195961713790894, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0232, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.1461530476808548, + "learning_rate": 2.457441244940383e-05, + "loss": 0.022, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13729079067707062, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0196, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.15615087747573853, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0213, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.12907880544662476, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0198, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.11320316791534424, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0225, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.14982816576957703, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0225, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.12824060022830963, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0196, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.11940523982048035, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0204, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946357995271683, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.0218, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.13296495378017426, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0229, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.1383914053440094, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0207, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.11321834474802017, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0202, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.1276783049106598, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0193, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1406320035457611, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0191, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.12060298770666122, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0195, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.08627335727214813, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0191, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1429424285888672, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0197, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.15133187174797058, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0212, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.11892743408679962, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0231, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.11871744692325592, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0211, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.105750173330307, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.0234, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10969540476799011, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.022, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.09607745707035065, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0173, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.11796857416629791, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0193, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.114372618496418, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.02, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.1164821982383728, + "learning_rate": 2.449463406055895e-05, + "loss": 0.023, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.13561253249645233, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.018, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.14956215023994446, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0245, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.12785622477531433, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0222, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.16085495054721832, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.021, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10792174935340881, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0209, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.10508404672145844, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0217, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.11471311002969742, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0202, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.14265185594558716, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0224, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.13569262623786926, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0196, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15635624527931213, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0205, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12042741477489471, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0179, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.09488346427679062, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0198, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.20503900945186615, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0199, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.13667535781860352, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13254515826702118, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0204, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.13879507780075073, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0233, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.11448678374290466, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0237, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1255042403936386, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0204, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11896912008523941, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0207, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.12624827027320862, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0194, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.09312308579683304, + "learning_rate": 2.442727178914008e-05, + "loss": 0.022, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12632645666599274, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0189, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.11468527466058731, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0182, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.09245705604553223, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0188, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.12113099545240402, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0191, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.13049040734767914, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0221, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.16979938745498657, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0211, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.11175577342510223, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0172, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.1191241443157196, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0206, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.16317729651927948, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.0213, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.10087717324495316, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0194, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.12195733189582825, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0253, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11880146712064743, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0222, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.14213789999485016, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0198, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.15624478459358215, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.0205, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09695806354284286, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0219, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11260177940130234, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0206, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.15394112467765808, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0219, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.11663617193698883, + "learning_rate": 2.436928131964383e-05, + "loss": 0.0214, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.12596042454242706, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0225, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.10310690104961395, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0203, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.12332364916801453, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0203, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.16257967054843903, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0234, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.13102053105831146, + "learning_rate": 2.435313188161029e-05, + "loss": 0.022, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.14736537635326385, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0199, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.14611469209194183, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0238, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.14594200253486633, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0212, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.16921328008174896, + "learning_rate": 2.434019954415699e-05, + "loss": 0.0236, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.1318638175725937, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0201, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.10869060456752777, + "learning_rate": 2.433372911848899e-05, + "loss": 0.019, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.1446760594844818, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.021, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.1298162341117859, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10307328402996063, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0197, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.11932921409606934, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0209, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.10986137390136719, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0227, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.12099824845790863, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0182, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.13432320952415466, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.0213, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.14018790423870087, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0199, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.13015322387218475, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0191, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.10281360149383545, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0215, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.09500012546777725, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0216, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.15131708979606628, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0225, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.08750247210264206, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0221, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.10143480449914932, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0214, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.12935201823711395, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0202, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.10703578591346741, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0207, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08649012446403503, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0184, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.11186458170413971, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.021, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.11181459575891495, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0211, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.11756999045610428, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0216, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.12061341106891632, + "learning_rate": 2.426561884159603e-05, + "loss": 0.019, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.13861820101737976, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0214, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.11545911431312561, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0195, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.14892302453517914, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0226, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.10354511439800262, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0197, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.15290679037570953, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0208, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.12185964733362198, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0205, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.10486926138401031, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0206, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.10959216207265854, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0198, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.11638635396957397, + "learning_rate": 2.423633359536448e-05, + "loss": 0.0227, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.13661672174930573, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0206, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.13071168959140778, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0189, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.16377446055412292, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0203, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.22221329808235168, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0185, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.11251460015773773, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0191, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.10873523354530334, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0228, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.13705629110336304, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0203, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.13980528712272644, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.10355913639068604, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0199, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.1282302737236023, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.0215, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.1221374049782753, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0192, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.18542620539665222, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0199, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1250954121351242, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0189, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12665164470672607, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0217, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.09464617818593979, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.0197, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.09657798707485199, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0194, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.11533467471599579, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0189, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.13647402822971344, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0194, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.13276426494121552, + "learning_rate": 2.41743228236291e-05, + "loss": 0.022, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.16899874806404114, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0209, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.14313359558582306, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0211, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.14507170021533966, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0218, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.13025349378585815, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0209, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.1666184365749359, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0217, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14891882240772247, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0205, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11392978578805923, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0223, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.12964752316474915, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.019, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11125730723142624, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0211, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.08378217369318008, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0203, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.14530996978282928, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0206, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.10187243670225143, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0188, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.09652446955442429, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0192, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.10552513599395752, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0212, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11044801771640778, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0215, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.12248426675796509, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0185, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.11164992302656174, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0198, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.11919678002595901, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0187, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1675145924091339, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0262, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.15451745688915253, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0226, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.0997307077050209, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0196, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.1272345781326294, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0214, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.1286081224679947, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0214, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12226564437150955, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0211, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.11795010417699814, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0216, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.12963736057281494, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0215, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.10415369272232056, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0215, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.11642655730247498, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0194, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.19006073474884033, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0216, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.16677096486091614, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.0235, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.15928123891353607, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0195, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.10270075500011444, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0197, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.12253808975219727, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0216, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14145877957344055, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0211, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1387409120798111, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0185, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.11433693021535873, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0204, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.14247572422027588, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0187, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.12203793972730637, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.018, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.15705470740795135, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0212, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.10648002475500107, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0182, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.11573450267314911, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0211, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.11960671842098236, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0245, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.14416365325450897, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0178, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.11234421283006668, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0205, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.1375260204076767, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.025, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.08686859905719757, + "learning_rate": 2.402315322351119e-05, + "loss": 0.018, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11352542787790298, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0198, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.13374295830726624, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0216, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.09358088672161102, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0202, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.13767337799072266, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0229, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.1405765861272812, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.02, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11317892372608185, + "learning_rate": 2.400332803040664e-05, + "loss": 0.02, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.12356667965650558, + "learning_rate": 2.400002143698994e-05, + "loss": 0.0225, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.12570610642433167, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0232, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.1461946964263916, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0201, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.10893318802118301, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0211, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.10969080775976181, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0235, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.17320629954338074, + "learning_rate": 2.398347822254704e-05, + "loss": 0.0221, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14296430349349976, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0206, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12349261343479156, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0182, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.11617975682020187, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0209, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.1574973613023758, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0222, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.157145157456398, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.0199, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.14698748290538788, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0203, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.12345635890960693, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0212, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.1370248794555664, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0194, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.1332298070192337, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0219, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.12525111436843872, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.0219, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.14905311167240143, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0211, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12063094973564148, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.0262, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.10361988842487335, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0195, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.09919667989015579, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0201, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.11141914129257202, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.0223, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.17785966396331787, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0201, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.126350998878479, + "learning_rate": 2.392710390356112e-05, + "loss": 0.0201, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.12107345461845398, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0213, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.11466097086668015, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0202, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.15397872030735016, + "learning_rate": 2.391713512040237e-05, + "loss": 0.022, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.14412778615951538, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0223, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.10495682060718536, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0194, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.13021183013916016, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0199, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.13993357121944427, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0182, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.128571018576622, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.0181, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.11090676486492157, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0204, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.10988310724496841, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0203, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.11705116927623749, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0214, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.1277252584695816, + "learning_rate": 2.388719222953459e-05, + "loss": 0.0225, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1561414748430252, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0221, + "step": 40000 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 20000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 64, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-60000/config.json b/checkpoint-60000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d7b77e4255263f3803a4b6f94ca6831e459b70 --- /dev/null +++ b/checkpoint-60000/config.json @@ -0,0 +1,64 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/checkpoint-60000/experiment_cfg/metadata.json b/checkpoint-60000/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..300eb070963337016e16465f50ba8351ea2eb25c --- /dev/null +++ b/checkpoint-60000/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.3139495849609375, + 0.4876587688922882, + 0.7196521759033203 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6996827125549316 + ], + "mean": [ + 2.749288365724937, + -2.0455216239909983, + 0.700532551020533 + ], + "std": [ + 1.6786295897046262, + 1.312897969434244, + 0.00133751758906258 + ], + "q01": [ + -1.4241907881148037, + -5.1716547935950885, + 0.6999670898768614 + ], + "q99": [ + 6.198111545831555, + -0.5873531334104489, + 0.7038833014242587 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.23084999587450247, + 0.6238431088581847 + ], + "std": [ + 0.0, + 0.0, + 0.6556404371644047, + 0.3572252105732042 + ], + "q01": [ + 0.0, + 0.0, + -0.9999999999999999, + 1.7482354593273349e-06 + ], + "q99": [ + 0.0, + 0.0, + 0.9999999999999999, + 0.9999999999999999 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8003877401351929, + 0.9829942584037781 + ], + "min": [ + -0.37471598386764526, + -0.8472502827644348, + -0.25070279836654663 + ], + "mean": [ + 0.28667332601863704, + -0.038315473170876704, + 0.45974658906777444 + ], + "std": [ + 0.17067058828305154, + 0.23569952037784195, + 0.21950702354181487 + ], + "q01": [ + 0.015172948963784228, + -0.44450502063220615, + 0.23314444103329338 + ], + "q99": [ + 0.5609069689572412, + 0.38454179128009547, + 0.7028102736473852 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9984455108642578, + 0.9434727430343628, + 0.9062229990959167 + ], + "min": [ + -0.9999930262565613, + -0.9988337755203247, + -0.9618149995803833, + 2.1872274658107926e-07 + ], + "mean": [ + -0.25672297401336, + 0.024425335806687216, + -0.08740977164483847, + 0.16608739644018397 + ], + "std": [ + 0.7932585208392383, + 0.3102260195580416, + 0.3772472576315148, + 0.17451959300158773 + ], + "q01": [ + -0.99379006987882, + -0.5478102050038828, + -0.6101777952971636, + 0.002274085014134748 + ], + "q99": [ + 0.8804856038896288, + 0.5910611092873037, + 0.5216058042993562, + 0.5231965974012255 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011436971835792065, + -0.05664053186774254 + ], + "mean": [ + 0.031651551516052555, + -0.03161481387920421 + ], + "std": [ + 0.013119769222217526, + 0.01306078225192402 + ], + "q01": [ + 0.0060999246446777336, + -0.04062601653964198 + ], + "q99": [ + 0.04054891621517283, + -0.0059163567113456345 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.008144896753718373, + -0.00018893135146078263, + -0.0008739062727221845, + 0.0 + ], + "std": [ + 0.11030045424364411, + 0.10148082570313594, + 0.08975698373467506, + 0.0 + ], + "q01": [ + -0.07287373067581518, + -0.09446994199569948, + -0.07702818259249572, + 0.0 + ], + "q99": [ + 0.04485485414407898, + 0.09626941605965177, + 0.07610665906122742, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9221974313425939 + ], + "std": [ + 0.38670194342612674 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.384693946883099 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.01976129923017491, + -0.020870631344490034, + -0.05993476517349181 + ], + "std": [ + 0.4347024990804053, + 0.4221662976569997, + 0.3845506843044066 + ], + "q01": [ + -0.8835792939767807, + -0.9280919251713778, + -0.783361071470956 + ], + "q99": [ + 0.7622084438449307, + 0.8625125252609077, + 0.7470974753250706 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.008089673068544335, + -0.026498254627927348, + 0.0029083551743059005 + ], + "std": [ + 0.11216734503733773, + 0.13206002613798629, + 0.12615870412692864 + ], + "q01": [ + -0.2814627972846672, + -0.4342386516115439, + -0.34489755628340574 + ], + "q99": [ + 0.31879353266939386, + 0.28411797683220563, + 0.3597718671669894 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.3824228874769045 + ], + "std": [ + 0.9240015096469786 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.5597272305813592 + ] + } + } + }, + "modalities": { + "video": { + "robot0_eye_in_hand": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_left": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_right": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": true, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + }, + "end_effector_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": true, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/checkpoint-60000/model-00001-of-00002.safetensors b/checkpoint-60000/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c9509993e12e3bc3b291d1a0398318e256c8a2b7 --- /dev/null +++ b/checkpoint-60000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:98cf98087479b3e570d7fc66a0d397f42423009d07149e2cac936dbf1290a117 +size 4999367032 diff --git a/checkpoint-60000/model-00002-of-00002.safetensors b/checkpoint-60000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6c9b207e2680823b67811037f1ef92a9f5fedbc1 --- /dev/null +++ b/checkpoint-60000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c966ae82934e4f401e1138f03ea5ab851e33e7b9f0ac68c5e0765ae0dd1903a +size 2586705312 diff --git a/checkpoint-60000/model.safetensors.index.json b/checkpoint-60000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/checkpoint-60000/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-60000/optimizer.pt b/checkpoint-60000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..b373846637039fb0f5f4eb1dab429a3cd2ac45e3 --- /dev/null +++ b/checkpoint-60000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:89c25be79cf115d0c0d2fcacf013fe085bcd9289fa05371967586010d923a4e6 +size 8550719959 diff --git a/checkpoint-60000/rng_state_0.pth b/checkpoint-60000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..f163512a5774cc284a90bff740bbf80b6c5ddbe8 --- /dev/null +++ b/checkpoint-60000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:052b85b5f8e8c58c246c09847a74a23f6a2ba15ca27a5712b4616b2271ed22e4 +size 14917 diff --git a/checkpoint-60000/rng_state_1.pth b/checkpoint-60000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..33cc2b82991952634eb9778fe65da57fb20b8b3c --- /dev/null +++ b/checkpoint-60000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e52166b089e12778dad9fb778e9d234445b9262f60a0b6ac58971955380ebc5 +size 14917 diff --git a/checkpoint-60000/scheduler.pt b/checkpoint-60000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..f07ddb3979d54ae433e2bed05c09c54f792ab61c --- /dev/null +++ b/checkpoint-60000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5d866d35e39d601da48396b18689a9a3534395223867c3d1da65c34ae9a7a30 +size 1465 diff --git a/checkpoint-60000/trainer_state.json b/checkpoint-60000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..40c41a2abedb5a73e0e750c198075b261775721c --- /dev/null +++ b/checkpoint-60000/trainer_state.json @@ -0,0 +1,42034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 17.457084666860634, + "eval_steps": 500, + "global_step": 60000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002909514111143439, + "grad_norm": 5.5908684730529785, + "learning_rate": 4.5e-08, + "loss": 0.5668, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.084368705749512, + "learning_rate": 9.5e-08, + "loss": 0.5781, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 5.755157947540283, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5759, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.174816608428955, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5202, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 6.199656963348389, + "learning_rate": 2.45e-07, + "loss": 0.5455, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.251872539520264, + "learning_rate": 2.95e-07, + "loss": 0.5507, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 5.606583595275879, + "learning_rate": 3.45e-07, + "loss": 0.5264, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.967868328094482, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.5166, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.92909574508667, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4578, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.306433200836182, + "learning_rate": 4.95e-07, + "loss": 0.4091, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.75846529006958, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4072, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.2030856609344482, + "learning_rate": 5.95e-07, + "loss": 0.2992, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.5756216049194336, + "learning_rate": 6.45e-07, + "loss": 0.2703, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.3420923948287964, + "learning_rate": 6.95e-07, + "loss": 0.2492, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 0.9191234111785889, + "learning_rate": 7.45e-07, + "loss": 0.2302, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.662022054195404, + "learning_rate": 7.95e-07, + "loss": 0.1993, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.704409658908844, + "learning_rate": 8.45e-07, + "loss": 0.1814, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.6104884147644043, + "learning_rate": 8.95e-07, + "loss": 0.1694, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.4955693781375885, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1674, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.707667887210846, + "learning_rate": 9.95e-07, + "loss": 0.1507, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.6977017521858215, + "learning_rate": 1.045e-06, + "loss": 0.1547, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.49085867404937744, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1536, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.6363996267318726, + "learning_rate": 1.145e-06, + "loss": 0.1565, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.4633055627346039, + "learning_rate": 1.195e-06, + "loss": 0.1641, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5155240297317505, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1513, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5815547108650208, + "learning_rate": 1.295e-06, + "loss": 0.1431, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.5865835547447205, + "learning_rate": 1.345e-06, + "loss": 0.1419, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.5872474312782288, + "learning_rate": 1.395e-06, + "loss": 0.1326, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.4897519052028656, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1326, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.46316635608673096, + "learning_rate": 1.495e-06, + "loss": 0.1442, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.5359047651290894, + "learning_rate": 1.545e-06, + "loss": 0.1394, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.5805708169937134, + "learning_rate": 1.595e-06, + "loss": 0.1231, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.5488415360450745, + "learning_rate": 1.645e-06, + "loss": 0.127, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.41755908727645874, + "learning_rate": 1.695e-06, + "loss": 0.1375, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.6258692145347595, + "learning_rate": 1.745e-06, + "loss": 0.1202, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.5017617344856262, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1317, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.710078239440918, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.46977895498275757, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1215, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.4802077114582062, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.116, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.6448793411254883, + "learning_rate": 1.995e-06, + "loss": 0.1172, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4394509494304657, + "learning_rate": 2.045e-06, + "loss": 0.1145, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.6547806859016418, + "learning_rate": 2.095e-06, + "loss": 0.1174, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.6954914331436157, + "learning_rate": 2.145e-06, + "loss": 0.1245, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.670100212097168, + "learning_rate": 2.195e-06, + "loss": 0.1133, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.5230806469917297, + "learning_rate": 2.245e-06, + "loss": 0.11, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.46809589862823486, + "learning_rate": 2.295e-06, + "loss": 0.1069, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.6178835034370422, + "learning_rate": 2.345e-06, + "loss": 0.1099, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.41567298769950867, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1072, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4968737065792084, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1093, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.46893924474716187, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1046, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.5041665434837341, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1014, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5833673477172852, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1095, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.5565881133079529, + "learning_rate": 2.645e-06, + "loss": 0.1057, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6213339567184448, + "learning_rate": 2.695e-06, + "loss": 0.1087, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.5638794898986816, + "learning_rate": 2.745e-06, + "loss": 0.099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.5594047904014587, + "learning_rate": 2.795e-06, + "loss": 0.1065, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5521039366722107, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1072, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.5013270378112793, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1045, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.44355231523513794, + "learning_rate": 2.945e-06, + "loss": 0.0985, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.5467342734336853, + "learning_rate": 2.995e-06, + "loss": 0.101, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5065611600875854, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0939, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.5587425827980042, + "learning_rate": 3.095e-06, + "loss": 0.0969, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5561433434486389, + "learning_rate": 3.145e-06, + "loss": 0.1009, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.5576680302619934, + "learning_rate": 3.195e-06, + "loss": 0.1006, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.7426344752311707, + "learning_rate": 3.245e-06, + "loss": 0.0959, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.64913409948349, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.0962, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.5271337628364563, + "learning_rate": 3.345e-06, + "loss": 0.0927, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.6992810368537903, + "learning_rate": 3.395e-06, + "loss": 0.0924, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.559481680393219, + "learning_rate": 3.445e-06, + "loss": 0.0886, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7303935885429382, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0918, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.45654281973838806, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0934, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.41098082065582275, + "learning_rate": 3.595e-06, + "loss": 0.0911, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.43506675958633423, + "learning_rate": 3.645e-06, + "loss": 0.0887, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5131197571754456, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0983, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.4571124017238617, + "learning_rate": 3.745e-06, + "loss": 0.0928, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.6749465465545654, + "learning_rate": 3.795e-06, + "loss": 0.0926, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.46945369243621826, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1016, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.6481541395187378, + "learning_rate": 3.895e-06, + "loss": 0.0917, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.6769195795059204, + "learning_rate": 3.945e-06, + "loss": 0.0924, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5225585103034973, + "learning_rate": 3.995e-06, + "loss": 0.0909, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.642867922782898, + "learning_rate": 4.045e-06, + "loss": 0.0829, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.4617709219455719, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0924, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.5350925326347351, + "learning_rate": 4.145e-06, + "loss": 0.0821, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.4871961176395416, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0896, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.5545480847358704, + "learning_rate": 4.245e-06, + "loss": 0.0891, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.6057930588722229, + "learning_rate": 4.295e-06, + "loss": 0.0876, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.515282154083252, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0873, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.5402254462242126, + "learning_rate": 4.395e-06, + "loss": 0.0908, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4669134318828583, + "learning_rate": 4.445e-06, + "loss": 0.0837, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.46819189190864563, + "learning_rate": 4.495e-06, + "loss": 0.0888, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35766035318374634, + "learning_rate": 4.545e-06, + "loss": 0.0829, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.6769008040428162, + "learning_rate": 4.595e-06, + "loss": 0.0817, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5884619355201721, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0892, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5613101720809937, + "learning_rate": 4.695e-06, + "loss": 0.0852, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.47087088227272034, + "learning_rate": 4.745e-06, + "loss": 0.0818, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.4887458384037018, + "learning_rate": 4.795e-06, + "loss": 0.0858, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.44256526231765747, + "learning_rate": 4.845e-06, + "loss": 0.0833, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5375688076019287, + "learning_rate": 4.895e-06, + "loss": 0.075, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.6297019124031067, + "learning_rate": 4.945e-06, + "loss": 0.0932, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.6576125621795654, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0829, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.724773108959198, + "learning_rate": 5.045e-06, + "loss": 0.0813, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.54575115442276, + "learning_rate": 5.095e-06, + "loss": 0.0873, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.5186223387718201, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0736, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.46651381254196167, + "learning_rate": 5.195e-06, + "loss": 0.0791, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.5323564410209656, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0753, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.42579638957977295, + "learning_rate": 5.295e-06, + "loss": 0.0755, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.5063076019287109, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0767, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.4558819532394409, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0807, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.6920697093009949, + "learning_rate": 5.445e-06, + "loss": 0.079, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.4706108868122101, + "learning_rate": 5.495e-06, + "loss": 0.079, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.46705493330955505, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0745, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4643767774105072, + "learning_rate": 5.595e-06, + "loss": 0.0778, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.4800759553909302, + "learning_rate": 5.645e-06, + "loss": 0.0792, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.37705323100090027, + "learning_rate": 5.695e-06, + "loss": 0.0815, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.6651142239570618, + "learning_rate": 5.745e-06, + "loss": 0.0732, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.5275036692619324, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0758, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.45047008991241455, + "learning_rate": 5.845e-06, + "loss": 0.0808, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.6931971311569214, + "learning_rate": 5.895e-06, + "loss": 0.0808, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.7053916454315186, + "learning_rate": 5.945e-06, + "loss": 0.0747, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.6798260807991028, + "learning_rate": 5.995e-06, + "loss": 0.0829, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.5724003314971924, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0749, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.5836488008499146, + "learning_rate": 6.095e-06, + "loss": 0.0759, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.40817850828170776, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0778, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.41070470213890076, + "learning_rate": 6.195e-06, + "loss": 0.0765, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.43642961978912354, + "learning_rate": 6.245e-06, + "loss": 0.0711, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.49526578187942505, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0713, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.7773705720901489, + "learning_rate": 6.345e-06, + "loss": 0.0744, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.5485936999320984, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0732, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.5593521595001221, + "learning_rate": 6.445e-06, + "loss": 0.0747, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.3667270839214325, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0723, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.5555917620658875, + "learning_rate": 6.545e-06, + "loss": 0.0726, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.38107141852378845, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0647, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.46452224254608154, + "learning_rate": 6.645e-06, + "loss": 0.0787, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6018183827400208, + "learning_rate": 6.695e-06, + "loss": 0.0718, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.44997578859329224, + "learning_rate": 6.745e-06, + "loss": 0.0758, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45717573165893555, + "learning_rate": 6.795e-06, + "loss": 0.0738, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.5185182690620422, + "learning_rate": 6.845e-06, + "loss": 0.0768, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.5135819315910339, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.43529245257377625, + "learning_rate": 6.945e-06, + "loss": 0.0691, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.49389752745628357, + "learning_rate": 6.995e-06, + "loss": 0.0678, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.4113156199455261, + "learning_rate": 7.045e-06, + "loss": 0.0651, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.4072611927986145, + "learning_rate": 7.095e-06, + "loss": 0.0691, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.4524228870868683, + "learning_rate": 7.145e-06, + "loss": 0.0688, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.6225605607032776, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0676, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4841763377189636, + "learning_rate": 7.245e-06, + "loss": 0.0678, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.46297937631607056, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0759, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.5585324168205261, + "learning_rate": 7.345e-06, + "loss": 0.0671, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.57328200340271, + "learning_rate": 7.395e-06, + "loss": 0.0717, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.5072467923164368, + "learning_rate": 7.445000000000001e-06, + "loss": 0.071, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.4630240797996521, + "learning_rate": 7.495e-06, + "loss": 0.0674, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.35908272862434387, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0704, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.5117342472076416, + "learning_rate": 7.595e-06, + "loss": 0.0675, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.5214630961418152, + "learning_rate": 7.645e-06, + "loss": 0.0702, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.39036089181900024, + "learning_rate": 7.695e-06, + "loss": 0.0709, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.558722734451294, + "learning_rate": 7.745e-06, + "loss": 0.0705, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.3612176179885864, + "learning_rate": 7.795e-06, + "loss": 0.0715, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.523444652557373, + "learning_rate": 7.845e-06, + "loss": 0.0685, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.5231291651725769, + "learning_rate": 7.895e-06, + "loss": 0.0717, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.50368332862854, + "learning_rate": 7.945e-06, + "loss": 0.0694, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.4126806855201721, + "learning_rate": 7.995e-06, + "loss": 0.0693, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.719352662563324, + "learning_rate": 8.045e-06, + "loss": 0.0651, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.4405677020549774, + "learning_rate": 8.095e-06, + "loss": 0.0645, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.34746286273002625, + "learning_rate": 8.145e-06, + "loss": 0.0687, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.43200597167015076, + "learning_rate": 8.195e-06, + "loss": 0.0655, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.3464098274707794, + "learning_rate": 8.245e-06, + "loss": 0.0639, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.27263233065605164, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0629, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.343181848526001, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3949901759624481, + "learning_rate": 8.395e-06, + "loss": 0.0643, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.4489574730396271, + "learning_rate": 8.445e-06, + "loss": 0.0698, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.4746992588043213, + "learning_rate": 8.495e-06, + "loss": 0.0702, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.3651655316352844, + "learning_rate": 8.545e-06, + "loss": 0.0673, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.471071720123291, + "learning_rate": 8.595e-06, + "loss": 0.0648, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.35623008012771606, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0698, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.41837432980537415, + "learning_rate": 8.695e-06, + "loss": 0.0639, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.45608940720558167, + "learning_rate": 8.745e-06, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.40346869826316833, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0625, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.479824960231781, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0703, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.38570791482925415, + "learning_rate": 8.895e-06, + "loss": 0.0643, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.36530473828315735, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0652, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37868866324424744, + "learning_rate": 8.995000000000001e-06, + "loss": 0.068, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.32546791434288025, + "learning_rate": 9.045e-06, + "loss": 0.0636, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4664970338344574, + "learning_rate": 9.095e-06, + "loss": 0.0671, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4767569601535797, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0625, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.41823142766952515, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0685, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.3709985613822937, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0681, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.3933296799659729, + "learning_rate": 9.295e-06, + "loss": 0.0678, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.3079732060432434, + "learning_rate": 9.345e-06, + "loss": 0.0598, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.33485978841781616, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0644, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.40312913060188293, + "learning_rate": 9.445e-06, + "loss": 0.0684, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.4334392547607422, + "learning_rate": 9.495e-06, + "loss": 0.0622, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.5787541270256042, + "learning_rate": 9.545e-06, + "loss": 0.0656, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.2827294170856476, + "learning_rate": 9.595e-06, + "loss": 0.0628, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4431678056716919, + "learning_rate": 9.645e-06, + "loss": 0.0633, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.3916102647781372, + "learning_rate": 9.695e-06, + "loss": 0.0632, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.41470348834991455, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0612, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.3224067986011505, + "learning_rate": 9.795e-06, + "loss": 0.061, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.4110186696052551, + "learning_rate": 9.845e-06, + "loss": 0.066, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.4786360561847687, + "learning_rate": 9.895e-06, + "loss": 0.0648, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3667179048061371, + "learning_rate": 9.945e-06, + "loss": 0.0663, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.31722185015678406, + "learning_rate": 9.995e-06, + "loss": 0.0588, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.29271745681762695, + "learning_rate": 1.0045e-05, + "loss": 0.0636, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.3536592423915863, + "learning_rate": 1.0095e-05, + "loss": 0.0625, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.3228510916233063, + "learning_rate": 1.0145e-05, + "loss": 0.067, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.38474392890930176, + "learning_rate": 1.0195e-05, + "loss": 0.0609, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.40410760045051575, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0638, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.3632320463657379, + "learning_rate": 1.0295e-05, + "loss": 0.0587, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.37775692343711853, + "learning_rate": 1.0345e-05, + "loss": 0.0673, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.3343006372451782, + "learning_rate": 1.0395e-05, + "loss": 0.0599, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3232511281967163, + "learning_rate": 1.0445e-05, + "loss": 0.0661, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.38730430603027344, + "learning_rate": 1.0495e-05, + "loss": 0.0649, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.38972100615501404, + "learning_rate": 1.0545e-05, + "loss": 0.0582, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.27242428064346313, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0583, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.30514073371887207, + "learning_rate": 1.0645e-05, + "loss": 0.065, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.45364341139793396, + "learning_rate": 1.0695e-05, + "loss": 0.0576, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.3863564729690552, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0579, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.33717092871665955, + "learning_rate": 1.0795e-05, + "loss": 0.0565, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38073790073394775, + "learning_rate": 1.0845e-05, + "loss": 0.0585, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4277617037296295, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0622, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.3545698821544647, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0609, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.30646640062332153, + "learning_rate": 1.0995e-05, + "loss": 0.0659, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.3607352375984192, + "learning_rate": 1.1045e-05, + "loss": 0.0621, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3624425530433655, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0626, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3856692612171173, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0604, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.4390200972557068, + "learning_rate": 1.1195e-05, + "loss": 0.0601, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.4108872711658478, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0618, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.32732832431793213, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0636, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.3602239787578583, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.29978248476982117, + "learning_rate": 1.1395e-05, + "loss": 0.058, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.48490920662879944, + "learning_rate": 1.1445e-05, + "loss": 0.0614, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.3278113901615143, + "learning_rate": 1.1495e-05, + "loss": 0.0611, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.37862712144851685, + "learning_rate": 1.1545e-05, + "loss": 0.0626, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.3514714241027832, + "learning_rate": 1.1595e-05, + "loss": 0.0583, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.27256545424461365, + "learning_rate": 1.1645e-05, + "loss": 0.062, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.3305833339691162, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0572, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.2881031930446625, + "learning_rate": 1.1745e-05, + "loss": 0.0641, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.3219377398490906, + "learning_rate": 1.1795e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.34189504384994507, + "learning_rate": 1.1845e-05, + "loss": 0.0607, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30541297793388367, + "learning_rate": 1.1895e-05, + "loss": 0.056, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.35898879170417786, + "learning_rate": 1.1945e-05, + "loss": 0.0598, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.33221277594566345, + "learning_rate": 1.1995e-05, + "loss": 0.0603, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.3435590863227844, + "learning_rate": 1.2045e-05, + "loss": 0.0553, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.24820426106452942, + "learning_rate": 1.2095e-05, + "loss": 0.0644, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.3749466836452484, + "learning_rate": 1.2145e-05, + "loss": 0.0615, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.3984697759151459, + "learning_rate": 1.2195e-05, + "loss": 0.0595, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3250048756599426, + "learning_rate": 1.2245e-05, + "loss": 0.0608, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.2794302701950073, + "learning_rate": 1.2295e-05, + "loss": 0.0578, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3811834752559662, + "learning_rate": 1.2345e-05, + "loss": 0.0612, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3160156309604645, + "learning_rate": 1.2395e-05, + "loss": 0.0541, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.3825189173221588, + "learning_rate": 1.2445e-05, + "loss": 0.0621, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.3966415226459503, + "learning_rate": 1.2495e-05, + "loss": 0.0598, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3615199625492096, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0616, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.440762996673584, + "learning_rate": 1.2595e-05, + "loss": 0.0599, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.345599889755249, + "learning_rate": 1.2645e-05, + "loss": 0.0597, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.2963278889656067, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0556, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.3720734417438507, + "learning_rate": 1.2745e-05, + "loss": 0.0618, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3937445878982544, + "learning_rate": 1.2795e-05, + "loss": 0.0624, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.4492146372795105, + "learning_rate": 1.2845e-05, + "loss": 0.0573, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.5172158479690552, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0594, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.3796320855617523, + "learning_rate": 1.2945e-05, + "loss": 0.0528, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.42423033714294434, + "learning_rate": 1.2995e-05, + "loss": 0.0615, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.4578545093536377, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0562, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.38345491886138916, + "learning_rate": 1.3095e-05, + "loss": 0.0574, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.2539816200733185, + "learning_rate": 1.3145e-05, + "loss": 0.0562, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.44323891401290894, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0588, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.4074377417564392, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0591, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.5252887606620789, + "learning_rate": 1.3295e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.39536044001579285, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0588, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.6300645470619202, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0584, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.312466561794281, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0588, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3810383379459381, + "learning_rate": 1.3495e-05, + "loss": 0.0597, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.30645400285720825, + "learning_rate": 1.3545e-05, + "loss": 0.0589, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.2857605218887329, + "learning_rate": 1.3595e-05, + "loss": 0.0589, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.3206622898578644, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0597, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3226073086261749, + "learning_rate": 1.3695e-05, + "loss": 0.0559, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.24912619590759277, + "learning_rate": 1.3745e-05, + "loss": 0.0547, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.29700762033462524, + "learning_rate": 1.3795e-05, + "loss": 0.0563, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.32627588510513306, + "learning_rate": 1.3845e-05, + "loss": 0.0602, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.48569780588150024, + "learning_rate": 1.3895e-05, + "loss": 0.0553, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.323607474565506, + "learning_rate": 1.3945e-05, + "loss": 0.0578, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.42626142501831055, + "learning_rate": 1.3995e-05, + "loss": 0.0539, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.3281888961791992, + "learning_rate": 1.4045e-05, + "loss": 0.0638, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.38216397166252136, + "learning_rate": 1.4095e-05, + "loss": 0.0572, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.2816051244735718, + "learning_rate": 1.4145e-05, + "loss": 0.0543, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.2604160010814667, + "learning_rate": 1.4195e-05, + "loss": 0.0555, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.34318771958351135, + "learning_rate": 1.4245e-05, + "loss": 0.0565, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.31248190999031067, + "learning_rate": 1.4295e-05, + "loss": 0.0583, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.3014426827430725, + "learning_rate": 1.4345e-05, + "loss": 0.0586, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3650360703468323, + "learning_rate": 1.4395e-05, + "loss": 0.0581, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.33519116044044495, + "learning_rate": 1.4445e-05, + "loss": 0.0611, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.3595008850097656, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.054, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.32799217104911804, + "learning_rate": 1.4545e-05, + "loss": 0.0574, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.29410457611083984, + "learning_rate": 1.4595e-05, + "loss": 0.0582, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.2727576792240143, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0588, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.36562401056289673, + "learning_rate": 1.4695e-05, + "loss": 0.0597, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.3351168930530548, + "learning_rate": 1.4745e-05, + "loss": 0.0558, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.23880337178707123, + "learning_rate": 1.4795e-05, + "loss": 0.0571, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.3910059928894043, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0573, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.3103587031364441, + "learning_rate": 1.4895e-05, + "loss": 0.0578, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.3265620768070221, + "learning_rate": 1.4945e-05, + "loss": 0.0542, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.3060731291770935, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0537, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.41589322686195374, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0546, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.3102424740791321, + "learning_rate": 1.5095e-05, + "loss": 0.0564, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.32286810874938965, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0519, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.3129369020462036, + "learning_rate": 1.5195e-05, + "loss": 0.0566, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.43735018372535706, + "learning_rate": 1.5245e-05, + "loss": 0.0552, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.440449595451355, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0589, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.3631616532802582, + "learning_rate": 1.5345e-05, + "loss": 0.0511, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.29124221205711365, + "learning_rate": 1.5395e-05, + "loss": 0.055, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.4247622489929199, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0562, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.36976534128189087, + "learning_rate": 1.5495e-05, + "loss": 0.0604, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.3580514192581177, + "learning_rate": 1.5545e-05, + "loss": 0.0547, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.2969493567943573, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0564, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28583595156669617, + "learning_rate": 1.5645e-05, + "loss": 0.0586, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.3220328390598297, + "learning_rate": 1.5695e-05, + "loss": 0.0588, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.32270652055740356, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0559, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.3170712888240814, + "learning_rate": 1.5795e-05, + "loss": 0.0566, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.2966325879096985, + "learning_rate": 1.5845e-05, + "loss": 0.0551, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.2788766622543335, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.051, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.3521708548069, + "learning_rate": 1.5945e-05, + "loss": 0.0558, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.35105040669441223, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0551, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.4221852123737335, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0573, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.4217511713504791, + "learning_rate": 1.6095e-05, + "loss": 0.0532, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.4519175589084625, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0489, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3427059054374695, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0544, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29204127192497253, + "learning_rate": 1.6245e-05, + "loss": 0.0525, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.24138635396957397, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0559, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.3599991500377655, + "learning_rate": 1.6345e-05, + "loss": 0.0544, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.3326258361339569, + "learning_rate": 1.6395e-05, + "loss": 0.0593, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31976261734962463, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0624, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.26379555463790894, + "learning_rate": 1.6495e-05, + "loss": 0.0529, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2576923966407776, + "learning_rate": 1.6545e-05, + "loss": 0.0527, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.30890268087387085, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0571, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.4547998309135437, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.3050350248813629, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0552, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.2984176278114319, + "learning_rate": 1.6745e-05, + "loss": 0.0528, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.34583792090415955, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0512, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.33085328340530396, + "learning_rate": 1.6845e-05, + "loss": 0.0563, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.33391326665878296, + "learning_rate": 1.6895e-05, + "loss": 0.0527, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.23657914996147156, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0509, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.3719659149646759, + "learning_rate": 1.6995e-05, + "loss": 0.0568, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.4268389940261841, + "learning_rate": 1.7045e-05, + "loss": 0.0525, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.2726958394050598, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0482, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27847009897232056, + "learning_rate": 1.7145e-05, + "loss": 0.0559, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.7195e-05, + "loss": 0.0588, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.2666028141975403, + "learning_rate": 1.7245e-05, + "loss": 0.0548, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.26048436760902405, + "learning_rate": 1.7295e-05, + "loss": 0.0592, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.22569318115711212, + "learning_rate": 1.7345e-05, + "loss": 0.0505, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.35350096225738525, + "learning_rate": 1.7395e-05, + "loss": 0.0566, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.3166314363479614, + "learning_rate": 1.7445e-05, + "loss": 0.0517, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.24918541312217712, + "learning_rate": 1.7495e-05, + "loss": 0.0545, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3416207730770111, + "learning_rate": 1.7545e-05, + "loss": 0.0544, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.22633257508277893, + "learning_rate": 1.7595e-05, + "loss": 0.0545, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.3972736597061157, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0591, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3213656544685364, + "learning_rate": 1.7695e-05, + "loss": 0.0537, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.30361926555633545, + "learning_rate": 1.7745e-05, + "loss": 0.0529, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.35521000623703003, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0549, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.4444983899593353, + "learning_rate": 1.7845e-05, + "loss": 0.0529, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.32495826482772827, + "learning_rate": 1.7895e-05, + "loss": 0.0557, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.2893896698951721, + "learning_rate": 1.7945e-05, + "loss": 0.056, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.2607886493206024, + "learning_rate": 1.7995e-05, + "loss": 0.055, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.3305937647819519, + "learning_rate": 1.8045e-05, + "loss": 0.0529, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.3003319799900055, + "learning_rate": 1.8095e-05, + "loss": 0.0515, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3523712456226349, + "learning_rate": 1.8145e-05, + "loss": 0.0524, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.27575406432151794, + "learning_rate": 1.8195e-05, + "loss": 0.0535, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.36445286870002747, + "learning_rate": 1.8245e-05, + "loss": 0.0552, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.41415491700172424, + "learning_rate": 1.8295e-05, + "loss": 0.0537, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.3738953471183777, + "learning_rate": 1.8345e-05, + "loss": 0.0555, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.3446105420589447, + "learning_rate": 1.8395e-05, + "loss": 0.0526, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.39016082882881165, + "learning_rate": 1.8445e-05, + "loss": 0.0536, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.4075721502304077, + "learning_rate": 1.8495e-05, + "loss": 0.0548, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.4071345925331116, + "learning_rate": 1.8545e-05, + "loss": 0.0539, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.34446755051612854, + "learning_rate": 1.8595e-05, + "loss": 0.057, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.3431609272956848, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0515, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.31537994742393494, + "learning_rate": 1.8695e-05, + "loss": 0.0534, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.4846970736980438, + "learning_rate": 1.8745e-05, + "loss": 0.0532, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.3109683394432068, + "learning_rate": 1.8795e-05, + "loss": 0.0547, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.28903284668922424, + "learning_rate": 1.8845e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.26592350006103516, + "learning_rate": 1.8895e-05, + "loss": 0.0499, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.27347302436828613, + "learning_rate": 1.8945e-05, + "loss": 0.052, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.3194405138492584, + "learning_rate": 1.8995e-05, + "loss": 0.0521, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.40457430481910706, + "learning_rate": 1.9045e-05, + "loss": 0.0523, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.3055919408798218, + "learning_rate": 1.9095e-05, + "loss": 0.0504, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.4911873936653137, + "learning_rate": 1.9145e-05, + "loss": 0.0531, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.30619072914123535, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0525, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.24404728412628174, + "learning_rate": 1.9245e-05, + "loss": 0.0498, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.31723177433013916, + "learning_rate": 1.9295e-05, + "loss": 0.0519, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.28105342388153076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0528, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.32965341210365295, + "learning_rate": 1.9395e-05, + "loss": 0.0504, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.2487417608499527, + "learning_rate": 1.9445e-05, + "loss": 0.0552, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3788810670375824, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0535, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.2561226189136505, + "learning_rate": 1.9545e-05, + "loss": 0.056, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.29964715242385864, + "learning_rate": 1.9595e-05, + "loss": 0.0565, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.386600136756897, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0468, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.37690475583076477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.35367539525032043, + "learning_rate": 1.9745e-05, + "loss": 0.0501, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.30189844965934753, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0524, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.27525416016578674, + "learning_rate": 1.9845e-05, + "loss": 0.0524, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.22557462751865387, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.3572654128074646, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0536, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.37130483984947205, + "learning_rate": 1.9995e-05, + "loss": 0.0507, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.40820422768592834, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0547, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.24222403764724731, + "learning_rate": 2.0095e-05, + "loss": 0.0536, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.35473307967185974, + "learning_rate": 2.0145e-05, + "loss": 0.0503, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24238434433937073, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0511, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.32721346616744995, + "learning_rate": 2.0245e-05, + "loss": 0.0474, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.3787407875061035, + "learning_rate": 2.0295e-05, + "loss": 0.0515, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.3435424566268921, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0521, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.250515341758728, + "learning_rate": 2.0395e-05, + "loss": 0.0539, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.2730806767940521, + "learning_rate": 2.0445e-05, + "loss": 0.0544, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.32704228162765503, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0516, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.26865822076797485, + "learning_rate": 2.0545e-05, + "loss": 0.0557, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.31514087319374084, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.051, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.2616795301437378, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0483, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2979176938533783, + "learning_rate": 2.0695e-05, + "loss": 0.056, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.2876877188682556, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0496, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.2801620662212372, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0533, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.3521801233291626, + "learning_rate": 2.0845e-05, + "loss": 0.051, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.3016152083873749, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0515, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.29837727546691895, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0483, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.255982369184494, + "learning_rate": 2.0995e-05, + "loss": 0.0482, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.20852255821228027, + "learning_rate": 2.1045e-05, + "loss": 0.05, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.2746782898902893, + "learning_rate": 2.1095e-05, + "loss": 0.0513, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.28745001554489136, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0486, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.3163246810436249, + "learning_rate": 2.1195e-05, + "loss": 0.0536, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.2279432713985443, + "learning_rate": 2.1245e-05, + "loss": 0.0497, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.2273164689540863, + "learning_rate": 2.1295e-05, + "loss": 0.0485, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.3124007284641266, + "learning_rate": 2.1345e-05, + "loss": 0.0473, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.2097717970609665, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0514, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.24428582191467285, + "learning_rate": 2.1445e-05, + "loss": 0.0543, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.47236987948417664, + "learning_rate": 2.1495e-05, + "loss": 0.0525, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.3367164433002472, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0525, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.34086376428604126, + "learning_rate": 2.1595e-05, + "loss": 0.049, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.3060646951198578, + "learning_rate": 2.1645e-05, + "loss": 0.0538, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.2997030019760132, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0553, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.2652549743652344, + "learning_rate": 2.1745e-05, + "loss": 0.051, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.3072361350059509, + "learning_rate": 2.1795e-05, + "loss": 0.0516, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2494215965270996, + "learning_rate": 2.1845e-05, + "loss": 0.0476, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.20796196162700653, + "learning_rate": 2.1895e-05, + "loss": 0.0553, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.24235095083713531, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.21033242344856262, + "learning_rate": 2.1995e-05, + "loss": 0.0542, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.26056164503097534, + "learning_rate": 2.2045e-05, + "loss": 0.0546, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.24798662960529327, + "learning_rate": 2.2095e-05, + "loss": 0.0529, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.22179707884788513, + "learning_rate": 2.2145e-05, + "loss": 0.0542, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.2660025358200073, + "learning_rate": 2.2195e-05, + "loss": 0.0496, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.3771500289440155, + "learning_rate": 2.2245e-05, + "loss": 0.0502, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.3059975206851959, + "learning_rate": 2.2295e-05, + "loss": 0.0482, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.2941029667854309, + "learning_rate": 2.2345e-05, + "loss": 0.0483, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3481833040714264, + "learning_rate": 2.2395e-05, + "loss": 0.0541, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.2729704678058624, + "learning_rate": 2.2445e-05, + "loss": 0.0514, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.44202038645744324, + "learning_rate": 2.2495e-05, + "loss": 0.0534, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.311088889837265, + "learning_rate": 2.2545e-05, + "loss": 0.0474, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30663055181503296, + "learning_rate": 2.2595e-05, + "loss": 0.0491, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.3530942499637604, + "learning_rate": 2.2645e-05, + "loss": 0.0503, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.27998727560043335, + "learning_rate": 2.2695e-05, + "loss": 0.0493, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.2679221034049988, + "learning_rate": 2.2745e-05, + "loss": 0.0512, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.348821222782135, + "learning_rate": 2.2795e-05, + "loss": 0.0489, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.260958194732666, + "learning_rate": 2.2845e-05, + "loss": 0.0474, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.2429468184709549, + "learning_rate": 2.2895e-05, + "loss": 0.053, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.2635575234889984, + "learning_rate": 2.2945e-05, + "loss": 0.053, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.3416514992713928, + "learning_rate": 2.2995e-05, + "loss": 0.0476, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.221172496676445, + "learning_rate": 2.3045e-05, + "loss": 0.0519, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.30541878938674927, + "learning_rate": 2.3095e-05, + "loss": 0.0522, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.25274574756622314, + "learning_rate": 2.3145e-05, + "loss": 0.0499, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.366253525018692, + "learning_rate": 2.3195e-05, + "loss": 0.0463, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.16825024783611298, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0493, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.3158467710018158, + "learning_rate": 2.3295e-05, + "loss": 0.0482, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.2887287139892578, + "learning_rate": 2.3345e-05, + "loss": 0.053, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.4717058539390564, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0515, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.3148113489151001, + "learning_rate": 2.3445e-05, + "loss": 0.0523, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.23815515637397766, + "learning_rate": 2.3495e-05, + "loss": 0.0523, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23897255957126617, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0484, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.2680213749408722, + "learning_rate": 2.3595e-05, + "loss": 0.0485, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.29999756813049316, + "learning_rate": 2.3645e-05, + "loss": 0.0506, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.30607151985168457, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.05, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.26077136397361755, + "learning_rate": 2.3745e-05, + "loss": 0.0558, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2638765275478363, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.049, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.24569877982139587, + "learning_rate": 2.3845e-05, + "loss": 0.053, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.29622989892959595, + "learning_rate": 2.3895e-05, + "loss": 0.0517, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.2598831355571747, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.05, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.25716227293014526, + "learning_rate": 2.3995e-05, + "loss": 0.0453, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.37305381894111633, + "learning_rate": 2.4045e-05, + "loss": 0.0539, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.3151630461215973, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0516, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.296941339969635, + "learning_rate": 2.4145e-05, + "loss": 0.0509, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.31995704770088196, + "learning_rate": 2.4195e-05, + "loss": 0.0498, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2426208257675171, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0475, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.23927678167819977, + "learning_rate": 2.4295e-05, + "loss": 0.05, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.1887904703617096, + "learning_rate": 2.4345e-05, + "loss": 0.05, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.18480589985847473, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0496, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.24515867233276367, + "learning_rate": 2.4445e-05, + "loss": 0.0464, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.25505101680755615, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0481, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.24435125291347504, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0502, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.25899314880371094, + "learning_rate": 2.4595e-05, + "loss": 0.0535, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.30518922209739685, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0481, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.26098036766052246, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0477, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.24723340570926666, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.26713746786117554, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0506, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.21166373789310455, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0502, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2602159082889557, + "learning_rate": 2.4895e-05, + "loss": 0.0506, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27918240427970886, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.049, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.2131885290145874, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0477, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.28778040409088135, + "learning_rate": 2.5045e-05, + "loss": 0.0487, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.3056010603904724, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.049, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.2850024700164795, + "learning_rate": 2.5145e-05, + "loss": 0.0475, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.20490600168704987, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0497, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.32371339201927185, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0491, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.32466208934783936, + "learning_rate": 2.5295e-05, + "loss": 0.0485, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.2779853343963623, + "learning_rate": 2.5345e-05, + "loss": 0.0485, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.3257012367248535, + "learning_rate": 2.5395e-05, + "loss": 0.0486, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2625649571418762, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0486, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.19846542179584503, + "learning_rate": 2.5495e-05, + "loss": 0.0494, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.2415613830089569, + "learning_rate": 2.5545e-05, + "loss": 0.0539, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.20438073575496674, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0476, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.2543872892856598, + "learning_rate": 2.5645e-05, + "loss": 0.0489, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.26969635486602783, + "learning_rate": 2.5695e-05, + "loss": 0.0557, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.27818888425827026, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0517, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.28826993703842163, + "learning_rate": 2.5795e-05, + "loss": 0.0499, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.19518250226974487, + "learning_rate": 2.5845e-05, + "loss": 0.0499, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.2639068067073822, + "learning_rate": 2.5895e-05, + "loss": 0.0459, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.26014742255210876, + "learning_rate": 2.5945e-05, + "loss": 0.0522, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.26659756898880005, + "learning_rate": 2.5995e-05, + "loss": 0.0471, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.24073593318462372, + "learning_rate": 2.6045e-05, + "loss": 0.0488, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.2724616527557373, + "learning_rate": 2.6095e-05, + "loss": 0.0493, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.3012109398841858, + "learning_rate": 2.6145e-05, + "loss": 0.049, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23306235671043396, + "learning_rate": 2.6195e-05, + "loss": 0.0474, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.23879121243953705, + "learning_rate": 2.6245e-05, + "loss": 0.048, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20701922476291656, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0527, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.28734415769577026, + "learning_rate": 2.6345e-05, + "loss": 0.0469, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.26227426528930664, + "learning_rate": 2.6395e-05, + "loss": 0.0472, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.24196277558803558, + "learning_rate": 2.6445e-05, + "loss": 0.0462, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.3405964970588684, + "learning_rate": 2.6495e-05, + "loss": 0.0479, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.24121852219104767, + "learning_rate": 2.6545e-05, + "loss": 0.0444, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.284856379032135, + "learning_rate": 2.6595e-05, + "loss": 0.0492, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.2515311539173126, + "learning_rate": 2.6645e-05, + "loss": 0.0453, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.29297393560409546, + "learning_rate": 2.6695e-05, + "loss": 0.049, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.2528100311756134, + "learning_rate": 2.6745e-05, + "loss": 0.0477, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.29126009345054626, + "learning_rate": 2.6795e-05, + "loss": 0.0484, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.26489511132240295, + "learning_rate": 2.6845e-05, + "loss": 0.0476, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.33150023221969604, + "learning_rate": 2.6895e-05, + "loss": 0.0475, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.25904157757759094, + "learning_rate": 2.6945e-05, + "loss": 0.0495, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.24830900132656097, + "learning_rate": 2.6995e-05, + "loss": 0.0476, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.277997225522995, + "learning_rate": 2.7045e-05, + "loss": 0.0478, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.3100328743457794, + "learning_rate": 2.7095e-05, + "loss": 0.0501, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.2957077622413635, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0491, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.19402435421943665, + "learning_rate": 2.7195e-05, + "loss": 0.0426, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.20080482959747314, + "learning_rate": 2.7245e-05, + "loss": 0.0464, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.21148671209812164, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0457, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.2299162596464157, + "learning_rate": 2.7345e-05, + "loss": 0.0467, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2296467125415802, + "learning_rate": 2.7395e-05, + "loss": 0.0489, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.25099772214889526, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0487, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2503034472465515, + "learning_rate": 2.7495e-05, + "loss": 0.0533, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.24678738415241241, + "learning_rate": 2.7545e-05, + "loss": 0.0462, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.3281477987766266, + "learning_rate": 2.7595e-05, + "loss": 0.0492, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.32214412093162537, + "learning_rate": 2.7645e-05, + "loss": 0.0459, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.29244086146354675, + "learning_rate": 2.7695e-05, + "loss": 0.0495, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.2372455894947052, + "learning_rate": 2.7745e-05, + "loss": 0.0477, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.28411489725112915, + "learning_rate": 2.7795e-05, + "loss": 0.0514, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.36217001080513, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2913915514945984, + "learning_rate": 2.7895e-05, + "loss": 0.0535, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.3237503170967102, + "learning_rate": 2.7945e-05, + "loss": 0.0509, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.32016751170158386, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0526, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.25484171509742737, + "learning_rate": 2.8045e-05, + "loss": 0.0508, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.22641469538211823, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.23269280791282654, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0484, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23346202075481415, + "learning_rate": 2.8195e-05, + "loss": 0.0472, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.25522616505622864, + "learning_rate": 2.8245e-05, + "loss": 0.0493, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.2478165179491043, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0468, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.25732970237731934, + "learning_rate": 2.8345e-05, + "loss": 0.0484, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.21047812700271606, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0455, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.252676397562027, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0496, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20048962533473969, + "learning_rate": 2.8495e-05, + "loss": 0.0434, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.1821785420179367, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0502, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.24106337130069733, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0471, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.413974404335022, + "learning_rate": 2.8645e-05, + "loss": 0.0542, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.29419073462486267, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0499, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.20694036781787872, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0453, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2390667200088501, + "learning_rate": 2.8795e-05, + "loss": 0.045, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2563699781894684, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.046, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.22678422927856445, + "learning_rate": 2.8895e-05, + "loss": 0.0456, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.26104822754859924, + "learning_rate": 2.8945e-05, + "loss": 0.0446, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.22175569832324982, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0478, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.20990405976772308, + "learning_rate": 2.9045e-05, + "loss": 0.0459, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.27277135848999023, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0477, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.22422750294208527, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.047, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.2740926742553711, + "learning_rate": 2.9195e-05, + "loss": 0.0444, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2977927327156067, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0477, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26431694626808167, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0488, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.22759316861629486, + "learning_rate": 2.9345e-05, + "loss": 0.0465, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.24671563506126404, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0501, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.24065586924552917, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0439, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.3334764838218689, + "learning_rate": 2.9495e-05, + "loss": 0.0475, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.23565605282783508, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.046, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.26102808117866516, + "learning_rate": 2.9595e-05, + "loss": 0.0494, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.34544840455055237, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.05, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.2476256936788559, + "learning_rate": 2.9695e-05, + "loss": 0.0477, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.2795628309249878, + "learning_rate": 2.9745e-05, + "loss": 0.0508, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.2756715714931488, + "learning_rate": 2.9795e-05, + "loss": 0.0484, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.20578329265117645, + "learning_rate": 2.9845e-05, + "loss": 0.0461, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.20354008674621582, + "learning_rate": 2.9895e-05, + "loss": 0.0431, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24710936844348907, + "learning_rate": 2.9945e-05, + "loss": 0.0459, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.32639119029045105, + "learning_rate": 2.9995e-05, + "loss": 0.045, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.27762553095817566, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0476, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.2903423607349396, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.3292105793952942, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0503, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.23503656685352325, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0446, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.24153585731983185, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.048, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.24157844483852386, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.046, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.23378467559814453, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0478, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.20753100514411926, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0527, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.2574090361595154, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0468, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.23341813683509827, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0445, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21378666162490845, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0487, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.2410328984260559, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0488, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.22005778551101685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0478, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.3197212219238281, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.046, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.253745436668396, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0437, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.1991635262966156, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0474, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.20773734152317047, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0439, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.3026197552680969, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0462, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.22027666866779327, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0443, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.2512618899345398, + "learning_rate": 2.99997744429684e-05, + "loss": 0.044, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.23001381754875183, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0479, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.20057332515716553, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0433, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.2831740975379944, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0462, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.18362601101398468, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0472, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.29741403460502625, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0443, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.20039376616477966, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0457, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.14999084174633026, + "learning_rate": 2.999958785135414e-05, + "loss": 0.043, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.18134453892707825, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0416, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.25010257959365845, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0494, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.31700536608695984, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0455, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.25080761313438416, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0436, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.2008715718984604, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0459, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.21441133320331573, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0492, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.23232029378414154, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0459, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.30173608660697937, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0505, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.21179363131523132, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0438, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.21905283629894257, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0439, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.17932383716106415, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0452, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.21132917702198029, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0454, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.22732721269130707, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.047, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.28214913606643677, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0446, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.23979637026786804, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0454, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.3095552921295166, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0449, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.2585015594959259, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0473, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.18226519227027893, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0472, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2568792402744293, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0443, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.27914026379585266, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0423, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.21452564001083374, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0457, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.22575126588344574, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0458, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.24053922295570374, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0453, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.21956735849380493, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.048, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2325354665517807, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0439, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.1692194789648056, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.044, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.25796884298324585, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0416, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.2719244956970215, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0445, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.23312833905220032, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0488, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.24458816647529602, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0473, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.24380528926849365, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0411, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.1836114227771759, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0433, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.21051307022571564, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0462, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.18093563616275787, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.045, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2444952130317688, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0485, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.19429947435855865, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0487, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.2294744998216629, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0479, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.2147439569234848, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0492, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.2777695059776306, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0419, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22944948077201843, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0458, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.2822294235229492, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0429, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.20201396942138672, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0435, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.1961182802915573, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0441, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.1653209924697876, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0473, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.2082156240940094, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0508, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.199684277176857, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0466, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.20871016383171082, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0436, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.22781118750572205, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0504, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2652275562286377, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0467, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.20812813937664032, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0456, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2533996105194092, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0458, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.18139654397964478, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0441, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.27243369817733765, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0472, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.326709121465683, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0487, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.26698657870292664, + "learning_rate": 2.999617967180403e-05, + "loss": 0.05, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.4119804799556732, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0455, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.21175070106983185, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0462, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.22503416240215302, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.0468, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.254734605550766, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0471, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26713427901268005, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0439, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.2739362120628357, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0455, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.26221534609794617, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0451, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26752719283103943, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.23052464425563812, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0481, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.162479966878891, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0435, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.24716691672801971, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0459, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.14981329441070557, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0407, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.23326952755451202, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0459, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.2200203239917755, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.046, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.21501530706882477, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0433, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.22950918972492218, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0454, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.2042679488658905, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0444, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.16468611359596252, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0434, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.2180141657590866, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0447, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.1658746749162674, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.24921023845672607, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0398, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0445, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.25189560651779175, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0429, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.2614912688732147, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0427, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.2890806794166565, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0458, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.24215207993984222, + "learning_rate": 2.999336925698801e-05, + "loss": 0.04, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.1806204915046692, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0435, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.18884161114692688, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0422, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.2760990262031555, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0427, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.22660709917545319, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0459, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.24862384796142578, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.045, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.1911306381225586, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0433, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.2612020969390869, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0414, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.20945195853710175, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.043, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.2436099350452423, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0472, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.23788926005363464, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0455, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.20128341019153595, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0456, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.24002164602279663, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0414, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.2653037905693054, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.042, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.15122903883457184, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0395, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.19299687445163727, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0452, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.2534617781639099, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0445, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.29057401418685913, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0398, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.19744732975959778, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.041, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.25267109274864197, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0439, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2406993955373764, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0445, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.2561663091182709, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0465, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.2137032002210617, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0407, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.3002893030643463, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0433, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.32190272212028503, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0444, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.2076684534549713, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0432, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.31036099791526794, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0431, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.21386218070983887, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0442, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.19012001156806946, + "learning_rate": 2.998948184763285e-05, + "loss": 0.048, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.2310018688440323, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0466, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.1984279453754425, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0443, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.24101053178310394, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0446, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.17791084945201874, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0403, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.2016495168209076, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0404, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.19435155391693115, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.21197110414505005, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0398, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.22241441905498505, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.20825718343257904, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0434, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.24533888697624207, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.046, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.2455352544784546, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0412, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.18937505781650543, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0425, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.22356615960597992, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0418, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.23797748982906342, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0457, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.23130296170711517, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0427, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.20548853278160095, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0439, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19323743879795074, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0432, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.24339382350444794, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0412, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.18780124187469482, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0416, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.19353246688842773, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0461, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.21522872149944305, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0412, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.26847484707832336, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.3781932294368744, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0401, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.3219687342643738, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0423, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.1699237823486328, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0421, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.2174658179283142, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0398, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.2433294802904129, + "learning_rate": 2.998488804893431e-05, + "loss": 0.042, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.2116415947675705, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0405, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.22813065350055695, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0432, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.24252736568450928, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.048, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.2009168565273285, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0431, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.19379755854606628, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.23313777148723602, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0384, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.22056345641613007, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0459, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.3403800129890442, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0439, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.17649835348129272, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0419, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.20594412088394165, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0418, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.2901265323162079, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0405, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.22897225618362427, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0424, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.20961393415927887, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0438, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.28585514426231384, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0477, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2239968478679657, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0398, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.1651894897222519, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0405, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.1731943041086197, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0407, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.1708967089653015, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0417, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.17396977543830872, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0359, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.21327704191207886, + "learning_rate": 2.998095036004704e-05, + "loss": 0.045, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.15585461258888245, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0403, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.21774765849113464, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0403, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.1889292299747467, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0408, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.1639491617679596, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0425, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.23745876550674438, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.043, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19481611251831055, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0453, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.21001650393009186, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0429, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.2114192098379135, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0419, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.2506379783153534, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0392, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.21439281105995178, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0422, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.24763159453868866, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0433, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.1719428300857544, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0422, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.25943970680236816, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0425, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22226977348327637, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0448, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.2018195390701294, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0408, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.20869500935077667, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0443, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.23311494290828705, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.042, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.2501373589038849, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0382, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.2650911509990692, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0377, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.21386376023292542, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0405, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.27908873558044434, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0431, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.18731699883937836, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0402, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.2566622197628021, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0419, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.2418505847454071, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0406, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.199744313955307, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0424, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.20134931802749634, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.042, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.15607263147830963, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0453, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.22614715993404388, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0413, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.22810052335262299, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0433, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.16557516157627106, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0424, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.20254060626029968, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0394, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2068638950586319, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0408, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.27261367440223694, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0433, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.23216570913791656, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0445, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.2546656131744385, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0381, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.21829853951931, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0472, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.17109332978725433, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0396, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.1746513843536377, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0413, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.22411789000034332, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0373, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.21182285249233246, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.038, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.19430142641067505, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0419, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.22196920216083527, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0395, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.20695635676383972, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0407, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.1721169799566269, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0396, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.23676829040050507, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0396, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.3019765317440033, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0443, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.23958736658096313, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0402, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.17959944903850555, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0434, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.1563226878643036, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0397, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.16746941208839417, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0402, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.26673364639282227, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0411, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.23634585738182068, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0427, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.19724313914775848, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0408, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.2072090059518814, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0389, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.18574072420597076, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0427, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.23390066623687744, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0411, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.2137259542942047, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0395, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.21458280086517334, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0422, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.1838904619216919, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.043, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.1834891140460968, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0389, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.2012249082326889, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.27410435676574707, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0473, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.21607685089111328, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0405, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.17525449395179749, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.043, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.23137639462947845, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0414, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.20958152413368225, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0404, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.21453802287578583, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.043, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.2522047758102417, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.04, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.23595386743545532, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0396, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.23375281691551208, + "learning_rate": 2.996358559911323e-05, + "loss": 0.042, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.21628527343273163, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0434, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.2463727742433548, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0391, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.19919127225875854, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.043, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.28369203209877014, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0463, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.30222833156585693, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0397, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.19382458925247192, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0402, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2075910121202469, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0392, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.2503066658973694, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0427, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.18662676215171814, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0397, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.2432677149772644, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0425, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.19900213181972504, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0422, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.12742096185684204, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.20504114031791687, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0404, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.19127227365970612, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0425, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.18699967861175537, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0419, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.20074906945228577, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0381, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.24741096794605255, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0389, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.22948721051216125, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0435, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.17772004008293152, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0412, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.228739395737648, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0406, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.18312685191631317, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0401, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.21340692043304443, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0403, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.17110474407672882, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0388, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.19640640914440155, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.2386539727449417, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.043, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.20800922811031342, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0412, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1889747679233551, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0377, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.1801699548959732, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0402, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.20975525677204132, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0426, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.22626695036888123, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0404, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.2192756086587906, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0438, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.21307197213172913, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0447, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.28529244661331177, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0412, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.26367515325546265, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.044, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.29029133915901184, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.14019586145877838, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0427, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19176369905471802, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0408, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.19753597676753998, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0395, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1900634765625, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0427, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.21698379516601562, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0452, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.21722441911697388, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0417, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.1760135143995285, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0405, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18030588328838348, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0425, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.20307210087776184, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0416, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.19713939726352692, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0401, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.18719582259655, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.19560874998569489, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.04, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1633717119693756, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0424, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.20183435082435608, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0397, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.23784901201725006, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0438, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.17358708381652832, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0393, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18991878628730774, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0387, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.20455898344516754, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.042, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.23264537751674652, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0447, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2586922347545624, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0402, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.20447152853012085, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0406, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.1814388930797577, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0412, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2679784595966339, + "learning_rate": 2.994497917870469e-05, + "loss": 0.039, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.1818714141845703, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0423, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.2153957635164261, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0391, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.1858970820903778, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0394, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.23248426616191864, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.17587599158287048, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0367, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.18720309436321259, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.039, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.17602671682834625, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0393, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.15901489555835724, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0418, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.18207870423793793, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0401, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.22800953686237335, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0416, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.20031091570854187, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0417, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.24054989218711853, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0438, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.201461061835289, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0369, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.15697605907917023, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0397, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.16894802451133728, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0403, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2567543089389801, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0421, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.20297718048095703, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0396, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.17331846058368683, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.038, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.2214743047952652, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0405, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.2107590287923813, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0388, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.1768934726715088, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0364, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.16018562018871307, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0405, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.18402984738349915, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0425, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.21482129395008087, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0399, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.20554907619953156, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0388, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.2401154637336731, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0369, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.15072989463806152, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0359, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.1649719625711441, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0399, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.2386736422777176, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0403, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.2389005720615387, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0416, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.14930105209350586, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0372, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.18874907493591309, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0356, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.21466751396656036, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0391, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1697709560394287, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.24686385691165924, + "learning_rate": 2.993190369330068e-05, + "loss": 0.04, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.19733406603336334, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0417, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.3981468677520752, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0375, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.20926636457443237, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0384, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.18989330530166626, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0365, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17171184718608856, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0386, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.20474061369895935, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0413, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.19314105808734894, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0358, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.1794378012418747, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0409, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.17263059318065643, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0385, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.17863675951957703, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0355, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.20071984827518463, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0386, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.2025180608034134, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0409, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.21840184926986694, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0397, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.197452113032341, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0379, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.15370187163352966, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0379, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.22221392393112183, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0396, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.18536874651908875, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.18744619190692902, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0378, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.21197383105754852, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0399, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.29221004247665405, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0355, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.2411249279975891, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.039, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.3187369108200073, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.036, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.18096214532852173, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0372, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.2743191719055176, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0401, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.1864418387413025, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0384, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.15518276393413544, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.039, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.18089881539344788, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0359, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.2218618243932724, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0427, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.17051172256469727, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0401, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.15884852409362793, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0378, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1561235636472702, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0383, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.19711007177829742, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0409, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.18595683574676514, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.043, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.18613699078559875, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0357, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.1390007734298706, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0356, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.25603482127189636, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0377, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.1902502328157425, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0381, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.18781402707099915, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0393, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.20504912734031677, + "learning_rate": 2.991569760600352e-05, + "loss": 0.042, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.17359918355941772, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0373, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.2029414027929306, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0412, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.23485742509365082, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0398, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1454097479581833, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0394, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.22343166172504425, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0386, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.17754614353179932, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0358, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.14276635646820068, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0405, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.2062550187110901, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0429, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2093406766653061, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.038, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.14510402083396912, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0374, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.22726939618587494, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0373, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.14905233681201935, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0392, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.15828309953212738, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0361, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.990945968675917e-05, + "loss": 0.037, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.1867590844631195, + "learning_rate": 2.990900562792633e-05, + "loss": 0.038, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.18278446793556213, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0358, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.24628379940986633, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0393, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.2195759266614914, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0372, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.2397889941930771, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0352, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.1563243269920349, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0361, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.16278347373008728, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0366, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1776551604270935, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0369, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.19893717765808105, + "learning_rate": 2.990533239951709e-05, + "loss": 0.039, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.21403104066848755, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0378, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.19885864853858948, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0393, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.18425336480140686, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0404, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.1930677592754364, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0369, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18271560966968536, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0351, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.16518554091453552, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0367, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.17990152537822723, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0349, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.19625109434127808, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0368, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.1891077160835266, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0407, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.20162540674209595, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.17586444318294525, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0346, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.13146457076072693, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0338, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.16059468686580658, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0366, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.1724517047405243, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0395, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.1847251057624817, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0365, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.20104378461837769, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.037, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.2712373733520508, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0347, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.18014875054359436, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0388, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.17479945719242096, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0388, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.24176514148712158, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0387, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.2453809529542923, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0381, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2105131894350052, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0369, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1763688176870346, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0384, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.16223029792308807, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0348, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.173785999417305, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.038, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.23269419372081757, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0393, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.19176329672336578, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0399, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.1559731364250183, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0375, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.15701691806316376, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.22872254252433777, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0371, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15535064041614532, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0378, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.21613062918186188, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0382, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.1504523903131485, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0386, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.177171990275383, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2307550013065338, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0354, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.2215310037136078, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0371, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.23132716119289398, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0389, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.1937142014503479, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0373, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.23848143219947815, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0355, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.23107734322547913, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0359, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1455046832561493, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0404, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.21890339255332947, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0421, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.1897345334291458, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0357, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.2016463577747345, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0363, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.15850050747394562, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0391, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.18169008195400238, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0377, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.26095741987228394, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0391, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.20983633399009705, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0355, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.21026623249053955, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.2104070633649826, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0343, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1801508367061615, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0369, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.22245049476623535, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.0358, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.19220919907093048, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0354, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.2285826951265335, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0391, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.15382888913154602, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0374, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.13551367819309235, + "learning_rate": 2.987759237035982e-05, + "loss": 0.039, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.23362696170806885, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0367, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.15972478687763214, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0319, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.23858226835727692, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0354, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.17988596856594086, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0391, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.21262454986572266, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0392, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.1443355828523636, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0384, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.24133095145225525, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0382, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.175334632396698, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.039, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.20340824127197266, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0376, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.16549751162528992, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0371, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.16527467966079712, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0373, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1418737769126892, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0378, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18754976987838745, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0355, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.13649867475032806, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0329, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.21408453583717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0348, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.20929333567619324, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.24083013832569122, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0378, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17924576997756958, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0369, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.22360439598560333, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0382, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.15963956713676453, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0361, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.19003787636756897, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0365, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2458433210849762, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0375, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.22048939764499664, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0382, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.19965334236621857, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0371, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.228081613779068, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0375, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.24077889323234558, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0347, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.17692111432552338, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0365, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.1996857225894928, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0358, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.24799935519695282, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0377, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.24300269782543182, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0362, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.2230309098958969, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0364, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.16640354692935944, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0346, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.14670157432556152, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.039, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18531887233257294, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0353, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.19566190242767334, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0391, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.1967019885778427, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0355, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.2161232978105545, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0375, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.17630350589752197, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0397, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18905608355998993, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0347, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.1890578716993332, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.036, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.15397298336029053, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0401, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0362, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.20983757078647614, + "learning_rate": 2.985388676461079e-05, + "loss": 0.036, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.1554982215166092, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0378, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.1961488276720047, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0356, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.17265456914901733, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0357, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.1275556981563568, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0353, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.1667783409357071, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0354, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.27920690178871155, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0385, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.12451875954866409, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0352, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.15794622898101807, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0339, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.16285695135593414, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0369, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.18990163505077362, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0376, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.1867877095937729, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0363, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.15270273387432098, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0394, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.1934751272201538, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0382, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16943520307540894, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0369, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.20546840131282806, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0336, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.18490619957447052, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0415, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.16439616680145264, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0384, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.13623777031898499, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0379, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.18209043145179749, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0357, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.19922122359275818, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0365, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.15637168288230896, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0323, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1781233251094818, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0329, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.17547845840454102, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0344, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.1770639270544052, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0358, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.2677673101425171, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0356, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.16740617156028748, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0344, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.13076575100421906, + "learning_rate": 2.983793562696423e-05, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.17586535215377808, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0381, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.1623464673757553, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.037, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.16550037264823914, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0373, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.20314498245716095, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0366, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.19387134909629822, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0377, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.17531387507915497, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0381, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.21466659009456635, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.037, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1693795919418335, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0339, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.21929262578487396, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0332, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.2253660261631012, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0339, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.2084304839372635, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0363, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.2513981759548187, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0357, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.19585898518562317, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0353, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.22262930870056152, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0357, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.19813011586666107, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0347, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.16626323759555817, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0346, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.271424800157547, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.036, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2283792346715927, + "learning_rate": 2.982684510938631e-05, + "loss": 0.04, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.15996390581130981, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0394, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.20269536972045898, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0387, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.20575514435768127, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0386, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.15178649127483368, + "learning_rate": 2.982433099619666e-05, + "loss": 0.037, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.23187878727912903, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0355, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.12988153100013733, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0369, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19367633759975433, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0367, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.15319304168224335, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0337, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.19311629235744476, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0383, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.14771778881549835, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0393, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.20381954312324524, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0414, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22292324900627136, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0358, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.13758333027362823, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0359, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.18837112188339233, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0339, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.1757460981607437, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0352, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.24256370961666107, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0346, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.2417164295911789, + "learning_rate": 2.981603574229741e-05, + "loss": 0.037, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.21129392087459564, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0371, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16878797113895416, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0367, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.17526096105575562, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0353, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.1812993884086609, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.037, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.16877815127372742, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0353, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.20823006331920624, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0384, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1536571979522705, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0356, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.1477498710155487, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0346, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.12916848063468933, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0318, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.15418170392513275, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0369, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.1883322149515152, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0369, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.1569332480430603, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0374, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.21140694618225098, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0365, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.26859667897224426, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0395, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.2182101309299469, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0382, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.18896763026714325, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0372, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.17398814857006073, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0338, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.14043423533439636, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0361, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.14995983242988586, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0363, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.15542328357696533, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0361, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.16333894431591034, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0337, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.18489807844161987, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0352, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.1794673502445221, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.037, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.1882123351097107, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0414, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.16483181715011597, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.04, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.1727719008922577, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0361, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.1968923956155777, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0389, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1838955134153366, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.034, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.23556186258792877, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0369, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.2518751323223114, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0331, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18682284653186798, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0371, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.2049458622932434, + "learning_rate": 2.979480675356209e-05, + "loss": 0.04, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.11846950650215149, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0347, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.21428292989730835, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0354, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.154558002948761, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0359, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1178995817899704, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0364, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.16926616430282593, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0373, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.1721247136592865, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0346, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.1975203901529312, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0351, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.17340779304504395, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0347, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.22424420714378357, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0384, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.1688208281993866, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.16542011499404907, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0343, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.15097306668758392, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0372, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.15530438721179962, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0345, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.16840992867946625, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.034, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.19959625601768494, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0375, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.15717248618602753, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0344, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.21473488211631775, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.04, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.18634898960590363, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0346, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.16032545268535614, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0377, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.19153323769569397, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0408, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.16018801927566528, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0334, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.1498652696609497, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0349, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.1652575582265854, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0344, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.15636402368545532, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0331, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.22963644564151764, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0359, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.16931135952472687, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0389, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.17646029591560364, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0322, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.19198739528656006, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0318, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.16826699674129486, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0389, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.1693607121706009, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.037, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.147409588098526, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0343, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.1984914094209671, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0329, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.18371893465518951, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0357, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.16279958188533783, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0377, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2122160941362381, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0367, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.20099543035030365, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0343, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.17249372601509094, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0325, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1106863021850586, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0336, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.2338913530111313, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0331, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.1545400768518448, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0321, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.17114387452602386, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0323, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.16218030452728271, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.037, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.14618389308452606, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0306, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.17804935574531555, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0348, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1512002944946289, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0318, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.14877542853355408, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.037, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.1693819761276245, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0325, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.17713536322116852, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0356, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.1719064563512802, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0358, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.19448526203632355, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0363, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.18047156929969788, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0326, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.20896470546722412, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0362, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.15999653935432434, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0318, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.17288383841514587, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0336, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.21483922004699707, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0367, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.19133612513542175, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0336, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.16010481119155884, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0377, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.1778886914253235, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0354, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.14997199177742004, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0326, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.16331784427165985, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0326, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.18247577548027039, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0367, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1838751882314682, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0338, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.17194764316082, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0365, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.1354181170463562, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0315, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.19473528861999512, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0345, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14786818623542786, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0354, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.1931883543729782, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0345, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.159915491938591, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0357, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14373569190502167, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0309, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.1468040943145752, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0339, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.1592605859041214, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0374, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.17073556780815125, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0365, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.13732466101646423, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0331, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.18718381226062775, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0337, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.179783895611763, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0383, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.15795794129371643, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0368, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.13824236392974854, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0343, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.2081076204776764, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0385, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.1407512128353119, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0338, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.1619846224784851, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0325, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.184702068567276, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0353, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.18360818922519684, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0349, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.17683662474155426, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0332, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.17690211534500122, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0364, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.18250274658203125, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0336, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.12902002036571503, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0344, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.1598006635904312, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0332, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.17691342532634735, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.037, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.2213180959224701, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0324, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.14762577414512634, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0321, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.14543993771076202, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.034, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.18970227241516113, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0377, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.15049827098846436, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0341, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.1485714316368103, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0361, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1893187314271927, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0354, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.2345002293586731, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0336, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.24119973182678223, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0325, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.18577414751052856, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0337, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.1941320151090622, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0355, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.17646151781082153, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0342, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.13801687955856323, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0351, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.15124225616455078, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0337, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19114945828914642, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.033, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.23505449295043945, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0357, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.15116260945796967, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.15656375885009766, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0322, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.18982963263988495, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0368, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.18300476670265198, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0379, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.16485247015953064, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0335, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.19399121403694153, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0353, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.22944431006908417, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0343, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.16755999624729156, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0351, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.15779320895671844, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0318, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.2113935351371765, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0343, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.14698673784732819, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0332, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19958600401878357, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0357, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20250289142131805, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0334, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.1955910474061966, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0381, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.13143742084503174, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0321, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.15840105712413788, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0333, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.20588421821594238, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0347, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.1881185919046402, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0345, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.16714425384998322, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.032, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.16526015102863312, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0316, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.1865321844816208, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0351, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.19202300906181335, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0325, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.15551044046878815, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0339, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.15684190392494202, + "learning_rate": 2.969840338753429e-05, + "loss": 0.034, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.16806578636169434, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0319, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.1554284542798996, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0348, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1836974322795868, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0375, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.20003411173820496, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0364, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.1747797578573227, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0328, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.2204974740743637, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0331, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.17183512449264526, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0381, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.14398148655891418, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.03, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.17428989708423615, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0329, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.13864420354366302, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0319, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.15052364766597748, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0357, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.21513856947422028, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0349, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19239531457424164, + "learning_rate": 2.968758727655112e-05, + "loss": 0.0377, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.22284342348575592, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0358, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.17206035554409027, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0349, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.1991468220949173, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0316, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.13542494177818298, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0329, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.14473386108875275, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0346, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.1972355991601944, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0348, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.17447993159294128, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0337, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.14419333636760712, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0355, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.19937965273857117, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0381, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.18988190591335297, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0331, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.1552465558052063, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0333, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.1769426017999649, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0344, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.17652767896652222, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0354, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1598699688911438, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0354, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.17556554079055786, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0372, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.18424467742443085, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0372, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.1591036170721054, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.038, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.1973133534193039, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0332, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.139867901802063, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0336, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.17536301910877228, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0339, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16841425001621246, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0319, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.24087557196617126, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.1419876366853714, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0341, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0315, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.14541317522525787, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0326, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.18411535024642944, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0382, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1769087016582489, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0355, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.16459472477436066, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0354, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.1770380288362503, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.034, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.16219645738601685, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0316, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.18393918871879578, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0366, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.1436525285243988, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0318, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.18282611668109894, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0376, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.16712592542171478, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0367, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.13408221304416656, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0305, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.17664852738380432, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0312, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.13730520009994507, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0323, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1278112679719925, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0322, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.1742715984582901, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0339, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.15022823214530945, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0348, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.17816156148910522, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0348, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.18016202747821808, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0357, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.21158602833747864, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0331, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1747511923313141, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0301, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.16840165853500366, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0363, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.1507093459367752, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0352, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.2081453651189804, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.035, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.1459163874387741, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0331, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.17468400299549103, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0339, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.15824441611766815, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0309, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.19502659142017365, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0305, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.1502636969089508, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.032, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.13870884478092194, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0356, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1596919149160385, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0347, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.1830880492925644, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0359, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.13567551970481873, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0319, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.12907959520816803, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0327, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.17357082664966583, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.0308, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.20002517104148865, + "learning_rate": 2.963613172841619e-05, + "loss": 0.035, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.13852699100971222, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0353, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.15445692837238312, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.11687827110290527, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0319, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.15454575419425964, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0366, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.1916663944721222, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0327, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.16530223190784454, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0309, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17331530153751373, + "learning_rate": 2.962976980219581e-05, + "loss": 0.032, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.17225238680839539, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0323, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.16390499472618103, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0324, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.16459178924560547, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0315, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.1416798084974289, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0348, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.1546057015657425, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0329, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15206973254680634, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0304, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.18791860342025757, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.03, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.17669418454170227, + "learning_rate": 2.962243236871436e-05, + "loss": 0.034, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.1335674524307251, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0323, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.17903365194797516, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0327, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.17771181464195251, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0341, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1532796174287796, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0328, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.18777944147586823, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0343, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.2553282082080841, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0342, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.24115385115146637, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0347, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.18773671984672546, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0324, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.18168678879737854, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0355, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.14390279352664948, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0323, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.13260799646377563, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0336, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.16469483077526093, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0341, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.1618199497461319, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0305, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.15524879097938538, + "learning_rate": 2.960942086834339e-05, + "loss": 0.034, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.14689622819423676, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0352, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.17949312925338745, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0329, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.2376648485660553, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0326, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.1302678883075714, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0356, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.16316285729408264, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0318, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.14314667880535126, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0347, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.13162434101104736, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0335, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.13411720097064972, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0334, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.20853494107723236, + "learning_rate": 2.960094148810064e-05, + "loss": 0.036, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.15634609758853912, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0314, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.16109681129455566, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0323, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.15969422459602356, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0292, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.13126519322395325, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0338, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.13577662408351898, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0323, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13226218521595, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0307, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.12283602356910706, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0305, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16819486021995544, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0316, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.1443309634923935, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.031, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.17339679598808289, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0335, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.12366446852684021, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0319, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.18073806166648865, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0355, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.16230227053165436, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.033, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.15816937386989594, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0331, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.15911494195461273, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0296, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.19500406086444855, + "learning_rate": 2.958564529914256e-05, + "loss": 0.034, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.16553588211536407, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0348, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.1780945360660553, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0297, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16839291155338287, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0317, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.26763999462127686, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0335, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.1292608082294464, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0319, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19598962366580963, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0346, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.16073159873485565, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0317, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.2037622481584549, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0324, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.11946277320384979, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0329, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.12799547612667084, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0329, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.23773977160453796, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0327, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.19241386651992798, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0371, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.15968935191631317, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0335, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.1712830811738968, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0302, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.16464142501354218, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0346, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.2122460901737213, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0319, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.16571341454982758, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.033, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1610543429851532, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0347, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14235250651836395, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0322, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.14999081194400787, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0299, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.033, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.1448807269334793, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0307, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.20123368501663208, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0315, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.16191548109054565, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.031, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.1475522518157959, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.17617955803871155, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0323, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.20633190870285034, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0287, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.13562145829200745, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0347, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.15800844132900238, + "learning_rate": 2.955719840324779e-05, + "loss": 0.034, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.1652759164571762, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0291, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.11819476634263992, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0309, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.17562547326087952, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0327, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.1934351772069931, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0342, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.1565367877483368, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.033, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.21048930287361145, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0357, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.16762706637382507, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0341, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.1267809122800827, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0322, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.18461033701896667, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0357, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.13451576232910156, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0327, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.15478897094726562, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0311, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1461619883775711, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0314, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.1432616114616394, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0339, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.14061163365840912, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0337, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.1950581669807434, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0315, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.13942912220954895, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.031, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.15901322662830353, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0324, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.19862298667430878, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0322, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.15155215561389923, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0302, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16394968330860138, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0299, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.13225089013576508, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0305, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.18203529715538025, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0319, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.15323573350906372, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0346, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.18822363018989563, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.038, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1510896533727646, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0336, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.12274178862571716, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.032, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.1287696361541748, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0341, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.16619601845741272, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0321, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.12426833063364029, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0298, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.1751766800880432, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0341, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.13578996062278748, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0306, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.19475996494293213, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0314, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.13291390240192413, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0341, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.16193236410617828, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0325, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.16108311712741852, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0356, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.16233862936496735, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0303, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.14042457938194275, + "learning_rate": 2.951955450025455e-05, + "loss": 0.031, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.14922280609607697, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0336, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.1389254927635193, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0324, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.21042059361934662, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0331, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.18529878556728363, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0311, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1345708668231964, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0316, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.14794935286045074, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0367, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.18609392642974854, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0326, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.1762595772743225, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0312, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.15549533069133759, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0342, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.1910371482372284, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0316, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.12789981067180634, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0286, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.11457879841327667, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0303, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.12534750998020172, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0313, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.17089471220970154, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0318, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.16774669289588928, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0326, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.18693485856056213, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0356, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.20419354736804962, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15851624310016632, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0328, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1284184604883194, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.1148369237780571, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0291, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.15302640199661255, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.037, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.16082480549812317, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.031, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.16006684303283691, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0346, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14866025745868683, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0323, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.15490789711475372, + "learning_rate": 2.949326462724779e-05, + "loss": 0.033, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.18960629403591156, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0302, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.16329076886177063, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0338, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1700190156698227, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0326, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.14541998505592346, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0332, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.13809162378311157, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.16794940829277039, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0345, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1897762268781662, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0304, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2064688503742218, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0332, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.14193041622638702, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0332, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.15571711957454681, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0339, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.14264710247516632, + "learning_rate": 2.948147907679037e-05, + "loss": 0.036, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.1411963850259781, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0344, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.18891720473766327, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0308, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.15993206202983856, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0339, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.15609434247016907, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0315, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.19272638857364655, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0325, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.2836095690727234, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0323, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16382543742656708, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0335, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.18899215757846832, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0301, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.17764721810817719, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0322, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.1808568239212036, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0304, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.166172593832016, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0298, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.18262767791748047, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0305, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.17854458093643188, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.033, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.16440987586975098, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.033, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.19478020071983337, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0288, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.2605016231536865, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0318, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1850014477968216, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0313, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.1339086890220642, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.17828164994716644, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0315, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.1235261932015419, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0328, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.13704872131347656, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0322, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1296597719192505, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0315, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.16906750202178955, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0347, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14643864333629608, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.031, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.16102853417396545, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0315, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.17575663328170776, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0312, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1488381028175354, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0318, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.14629733562469482, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0333, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.17051763832569122, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.034, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.14636048674583435, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0315, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.19957765936851501, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.032, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.15027561783790588, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0342, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1580146700143814, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.032, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1452188491821289, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0302, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.1924576759338379, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0323, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.199595108628273, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0326, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.1557701975107193, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0303, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.21232250332832336, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0324, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.13840343058109283, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0321, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.14219941198825836, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0314, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1747855842113495, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0342, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1730964183807373, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0334, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.12938249111175537, + "learning_rate": 2.94341324778745e-05, + "loss": 0.029, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1782408058643341, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0299, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.943188093053675e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.12131186574697495, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0277, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.19141671061515808, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0307, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.2431664764881134, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0326, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15145368874073029, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0329, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.19781085848808289, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0332, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.19477561116218567, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0317, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.1818031370639801, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0352, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.13725785911083221, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0272, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.1874770075082779, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0322, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.18701452016830444, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0304, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.15173715353012085, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.032, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.1315767467021942, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0331, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.16627946496009827, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0299, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.2122701108455658, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0308, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.12093837559223175, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0303, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.2072226107120514, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0331, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.14395366609096527, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0323, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.18454021215438843, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0349, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1647966057062149, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0342, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1907777190208435, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0347, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.14438189566135406, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0295, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.16802345216274261, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0283, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.1488419622182846, + "learning_rate": 2.940567320758981e-05, + "loss": 0.033, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14881044626235962, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0318, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.18762606382369995, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0322, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.15426765382289886, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0314, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.17882433533668518, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0361, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.16833186149597168, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0319, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.16899487376213074, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0283, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.15301142632961273, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0325, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.13270439207553864, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0305, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.16603541374206543, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0347, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.21959784626960754, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0299, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.24568438529968262, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0339, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.14889934659004211, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0284, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.20354624092578888, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0309, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.18124797940254211, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0352, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.13871392607688904, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0311, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.18997471034526825, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0299, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.09903991222381592, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0304, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.13336266577243805, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0306, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.1564595103263855, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0295, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.17495550215244293, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.17048844695091248, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0266, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.1257491409778595, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0299, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.15244479477405548, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0341, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.15737923979759216, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.03, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.13226579129695892, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0268, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.21549297869205475, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0323, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.2102872133255005, + "learning_rate": 2.937416924082395e-05, + "loss": 0.029, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.14369380474090576, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0306, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.1390914022922516, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0273, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.13536471128463745, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0299, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.13321688771247864, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0298, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13358351588249207, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0309, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.1732599437236786, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0319, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.17455634474754333, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0314, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.1714484691619873, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0345, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.18120428919792175, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0293, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.21681863069534302, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0335, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.12099924683570862, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0299, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.13233523070812225, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0372, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.16708864271640778, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0305, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.2122814804315567, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0319, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14159013330936432, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0343, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.13396364450454712, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0308, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.17630934715270996, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0349, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.16417403519153595, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0316, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.16593796014785767, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0291, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.13613900542259216, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0299, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.15980054438114166, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0319, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1578991562128067, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0288, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.1628173291683197, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0317, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.15937024354934692, + "learning_rate": 2.934549758921277e-05, + "loss": 0.031, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.15258124470710754, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0323, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.16142596304416656, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0289, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.15643659234046936, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0295, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.17128989100456238, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0313, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.1871197372674942, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0314, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.1448175013065338, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0318, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.21338209509849548, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0334, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15262271463871002, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0293, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.14302000403404236, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0307, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.13756762444972992, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0288, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.1390128880739212, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0286, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.16088609397411346, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0299, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.11852050572633743, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0284, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.17093200981616974, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0331, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.2325763702392578, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0269, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.19948336482048035, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0314, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.16918544471263885, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0311, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.14563684165477753, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0301, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.19645430147647858, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0312, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.18950732052326202, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0292, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.11659497767686844, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0307, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1314595341682434, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0326, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1457567662000656, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.16525784134864807, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0322, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.15977199375629425, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0334, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.17081087827682495, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0308, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.20844970643520355, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.029, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.22166086733341217, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1766587346792221, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0298, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1307239532470703, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0298, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0289, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.17264464497566223, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0317, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.15076857805252075, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0308, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14443998038768768, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0291, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.13115015625953674, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0329, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1444655954837799, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0294, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.16740760207176208, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0318, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.14201150834560394, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0314, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.14225733280181885, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0297, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.16758595407009125, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0307, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.16751281917095184, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.03, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.1855877935886383, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0333, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.17420977354049683, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0277, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.1568649560213089, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0312, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.15874658524990082, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0297, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.13727585971355438, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0334, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.14999760687351227, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0328, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.16789689660072327, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.14890174567699432, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0312, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.18001733720302582, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0322, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.1718987226486206, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.03, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.18727312982082367, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0297, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.1528918594121933, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0296, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.13372628390789032, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0288, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.14980314671993256, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0319, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.11368083208799362, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0312, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.14968374371528625, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0311, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.15593667328357697, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0287, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.1762111783027649, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0296, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.15493011474609375, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0272, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.14142747223377228, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.03, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.1307222545146942, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0294, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.1598598062992096, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.03, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.17966923117637634, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0337, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15533174574375153, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0276, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.1223452091217041, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0294, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.12733329832553864, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0289, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.15903618931770325, + "learning_rate": 2.926085774098786e-05, + "loss": 0.029, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1539735347032547, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0272, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.20487286150455475, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0312, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.19920656085014343, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0315, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.15179695188999176, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0305, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.16460831463336945, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0315, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.21031928062438965, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0283, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.13835182785987854, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0287, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.1597253829240799, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0336, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.11488606780767441, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0328, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.18707934021949768, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0328, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.18874959647655487, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0317, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.1182878389954567, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0316, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.11695742607116699, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0308, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.1597018986940384, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0324, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.17729975283145905, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0316, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.18250048160552979, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0309, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.20568659901618958, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0293, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.12907297909259796, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0281, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.1809365600347519, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0333, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.17379151284694672, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0327, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.16082972288131714, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0308, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.17415443062782288, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0299, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.1659587323665619, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0272, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.1538572907447815, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0309, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.16303570568561554, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0322, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1444631963968277, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0306, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.24458004534244537, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0285, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.15384158492088318, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0277, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1425933837890625, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0303, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.12485234439373016, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0279, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.14287105202674866, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0297, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.1621903032064438, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0268, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.145540252327919, + "learning_rate": 2.921797581830774e-05, + "loss": 0.03, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.1410767287015915, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0309, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.10953503102064133, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0299, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.15029482543468475, + "learning_rate": 2.92140191009304e-05, + "loss": 0.031, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.13262124359607697, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0299, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1503380835056305, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0329, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.18932904303073883, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.028, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.1473148614168167, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0315, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.18515919148921967, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0309, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.16122235357761383, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0279, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.1684419959783554, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0309, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.21400189399719238, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0304, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.12552157044410706, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0286, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.15383073687553406, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.029, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.1269458532333374, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0289, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.1839684545993805, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0316, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.13441112637519836, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0281, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.13045959174633026, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0296, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.13933542370796204, + "learning_rate": 2.919408984114327e-05, + "loss": 0.029, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.15143601596355438, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0296, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.16731251776218414, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.032, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.13245129585266113, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0268, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.17244738340377808, + "learning_rate": 2.918873439576597e-05, + "loss": 0.028, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.1709858626127243, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0293, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18727441132068634, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0302, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1482084095478058, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0297, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.1343366652727127, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0303, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.1456393450498581, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0299, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1569104641675949, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0281, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.19711613655090332, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0298, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.14027468860149384, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.031, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.16852174699306488, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0289, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.18446287512779236, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0308, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.16207733750343323, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0283, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.19747264683246613, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0301, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.15206509828567505, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0295, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.18873688578605652, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0286, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.144051656126976, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0335, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.14620175957679749, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0284, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.19520114362239838, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0304, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.16127106547355652, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0333, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.11467311531305313, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0303, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11777719110250473, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0311, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.12151405215263367, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0278, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1852089911699295, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0297, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.19555963575839996, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0321, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.1584533303976059, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0292, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.1764054298400879, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0306, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.16257385909557343, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0298, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.13896049559116364, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0296, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.16509538888931274, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0321, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.14912301301956177, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0285, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.18058481812477112, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0302, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.18734610080718994, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0289, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.19707795977592468, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0327, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.16215947270393372, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0302, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1710345447063446, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0332, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.19638675451278687, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0316, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.17807847261428833, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0285, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.13119569420814514, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0298, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.13414712250232697, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0301, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.1941734254360199, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0327, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.19624412059783936, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0291, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.17615875601768494, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0293, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.19011878967285156, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0316, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.15080299973487854, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0263, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.1676572859287262, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0318, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.14975300431251526, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0263, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17691564559936523, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.17775340378284454, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0298, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.14438258111476898, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.14729741215705872, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0289, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.20924119651317596, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0297, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.15471352636814117, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0289, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.1571032702922821, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0307, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15048687160015106, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0301, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.15063059329986572, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0279, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15528838336467743, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0299, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.1756942719221115, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0291, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.1491880565881729, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0296, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.1562451720237732, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0302, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.12366753816604614, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0257, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.15308785438537598, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.16614912450313568, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0301, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.12428635358810425, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0303, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.135500967502594, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0283, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.16944879293441772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0279, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.174363911151886, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0295, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.16406473517417908, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0284, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.2144644856452942, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0277, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.22006124258041382, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0264, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.15261462330818176, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0298, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.15205760300159454, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.14101643860340118, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0305, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16015839576721191, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.17772945761680603, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0312, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16818812489509583, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0283, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1448637843132019, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0325, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.14702047407627106, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0303, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.15547925233840942, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0299, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.18527188897132874, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.029, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.1533391922712326, + "learning_rate": 2.907944011202334e-05, + "loss": 0.029, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.15355326235294342, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0286, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.1503538340330124, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0312, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.12572433054447174, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0277, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.13614071905612946, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0303, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.13363786041736603, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0292, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.13318102061748505, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.0289, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.14687903225421906, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0301, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.13144652545452118, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0271, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.16351011395454407, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0292, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.19256997108459473, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0268, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13006120920181274, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0285, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.12932874262332916, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.11981740593910217, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.14302268624305725, + "learning_rate": 2.905937392537971e-05, + "loss": 0.029, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.1446913331747055, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0289, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.1500432789325714, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.15604284405708313, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0283, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.1795264482498169, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0289, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.14356553554534912, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0304, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.1323058009147644, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0302, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.2098175585269928, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0297, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.1424631029367447, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0295, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.1956447958946228, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0294, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.11993541568517685, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0273, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.18414588272571564, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0316, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.17320625483989716, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0319, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.17650462687015533, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0304, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.18740026652812958, + "learning_rate": 2.903909846652783e-05, + "loss": 0.034, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.12898528575897217, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0309, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.11721221357584, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0297, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.1659565567970276, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0294, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.1469241976737976, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0298, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14588896930217743, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0332, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.16243143379688263, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0305, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.15857404470443726, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0318, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.14992372691631317, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0305, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.13760966062545776, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0299, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.14302772283554077, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0313, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.18560320138931274, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0262, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.13117150962352753, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0286, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.16370882093906403, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0298, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.18689852952957153, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.031, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.16412676870822906, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0289, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.20153313875198364, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0297, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.15255704522132874, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0316, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.2036312222480774, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0306, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.16087041795253754, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0315, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.1514595001935959, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0295, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.13961558043956757, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0259, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1996241807937622, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0336, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.15389376878738403, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0278, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.12345903366804123, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.032, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.16360828280448914, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0292, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.19773486256599426, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.032, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.11931311339139938, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0268, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.1378420740365982, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0296, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.09076640754938126, + "learning_rate": 2.899643488777395e-05, + "loss": 0.026, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.1116260215640068, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0266, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.18670536577701569, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0314, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.15351636707782745, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0279, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.15310993790626526, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0298, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.15458840131759644, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0329, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.12034284323453903, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0266, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.14272604882717133, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0296, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.16335424780845642, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0266, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.16105882823467255, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0355, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.1455371230840683, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0261, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.21555180847644806, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0301, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.15278705954551697, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0292, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.14099596440792084, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0287, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.10848120599985123, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.028, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.1380046308040619, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0273, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.14006251096725464, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0328, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.16518674790859222, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0286, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.13295377790927887, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.031, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.1623251736164093, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0281, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.16029980778694153, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.03, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.15575948357582092, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0294, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.157670795917511, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0294, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1470523178577423, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0281, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.12192463129758835, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0248, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.1240374892950058, + "learning_rate": 2.895893835994484e-05, + "loss": 0.028, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1385754942893982, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0287, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.140727236866951, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0281, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.15758906304836273, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0294, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.19664499163627625, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0277, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.12567110359668732, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0279, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.13595536351203918, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0313, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.14262230694293976, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0317, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1474338322877884, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0282, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.18601401150226593, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0274, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.1392786204814911, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0269, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.12022347748279572, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0262, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.17340584099292755, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0277, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.1438349187374115, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0273, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.1408676654100418, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0277, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.14853249490261078, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0291, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.20401965081691742, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.031, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.18385861814022064, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.12695786356925964, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0267, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.14115597307682037, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0291, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.15790492296218872, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0294, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.15104074776172638, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0278, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13686735928058624, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0276, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.12029985338449478, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0305, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1351509988307953, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0293, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.10609307885169983, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0285, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15056312084197998, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0292, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.17062142491340637, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0281, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.15565407276153564, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0298, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.136429563164711, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0282, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.1649797409772873, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0267, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.19815988838672638, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.032, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15660062432289124, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.03, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.1840171217918396, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0342, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.14063017070293427, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0282, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.14935986697673798, + "learning_rate": 2.890533053632388e-05, + "loss": 0.029, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.14691074192523956, + "learning_rate": 2.89037798520224e-05, + "loss": 0.033, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.14992092549800873, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0262, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.16885656118392944, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0301, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1223178282380104, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0275, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.1200852021574974, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0277, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.1251501888036728, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0252, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.16905754804611206, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0277, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.13312722742557526, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0296, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.13198228180408478, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0269, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.14326612651348114, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0253, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.14861920475959778, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0262, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.19368910789489746, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0295, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.143910750746727, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0282, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.12314137071371078, + "learning_rate": 2.888352491214935e-05, + "loss": 0.027, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.16095006465911865, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0272, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.1414606273174286, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.029, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.15631091594696045, + "learning_rate": 2.887882538107974e-05, + "loss": 0.0291, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16476406157016754, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0283, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.2236901968717575, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0286, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.1318763792514801, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0274, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.15661245584487915, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.029, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.13406990468502045, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0287, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16861270368099213, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0272, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.18082283437252045, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0285, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.17517460882663727, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0277, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.17011325061321259, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0325, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14378181099891663, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0306, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1766553819179535, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0304, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.1666528880596161, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0294, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.15178363025188446, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0273, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.14275510609149933, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0291, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.1568918377161026, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0267, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.1370164304971695, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0268, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.10388749092817307, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0289, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11448585987091064, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0302, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.15148931741714478, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0292, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.14388935267925262, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0257, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.14465516805648804, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0272, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.13414210081100464, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0273, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12774589657783508, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.028, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.24874050915241241, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0282, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.1624344289302826, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0294, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.12817884981632233, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0324, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1509743630886078, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0269, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.31345340609550476, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0287, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.1523321270942688, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0259, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.1633133739233017, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0269, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.14278745651245117, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.0253, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.18361203372478485, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0282, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.10244551301002502, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0268, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.125647634267807, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0264, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.12238023430109024, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0264, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.16535502672195435, + "learning_rate": 2.882169202177678e-05, + "loss": 0.0329, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.19624854624271393, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0303, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.18285983800888062, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0266, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.13460052013397217, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.0293, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.1427820920944214, + "learning_rate": 2.881525979398615e-05, + "loss": 0.0244, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.14855900406837463, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0276, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.1433439403772354, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0299, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.1452850103378296, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0271, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.14403793215751648, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0264, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.13841332495212555, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0277, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.1426801085472107, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0296, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.14492706954479218, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0272, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.1315297782421112, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.0289, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.1513436734676361, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0284, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.15784604847431183, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.0273, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.12729156017303467, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0297, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.20208381116390228, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0284, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.1486770659685135, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0271, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.1463368982076645, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.0292, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.12362723797559738, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0252, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.18984590470790863, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0292, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.12962038815021515, + "learning_rate": 2.878773564250086e-05, + "loss": 0.0301, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.132072314620018, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0298, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.1482372134923935, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0267, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.15687556564807892, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0308, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.11656040698289871, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.028, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.16482318937778473, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0256, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.13332504034042358, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.027, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.16453887522220612, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0253, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.16902299225330353, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0283, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.12483953684568405, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0276, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.14033488929271698, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0277, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.18922871351242065, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0258, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.15026360750198364, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0298, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17492659389972687, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0275, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.14793051779270172, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0301, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.17575892806053162, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.027, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.1661873608827591, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0275, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.14529843628406525, + "learning_rate": 2.875990888327069e-05, + "loss": 0.0277, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.11084084957838058, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0278, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1433780938386917, + "learning_rate": 2.875661528053842e-05, + "loss": 0.026, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.13195206224918365, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0298, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.14103750884532928, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0309, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.17067542672157288, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0296, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.14078323543071747, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0288, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.17906904220581055, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0268, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.163010373711586, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0285, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.17636623978614807, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0276, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.18726587295532227, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0297, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.18583180010318756, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.16773444414138794, + "learning_rate": 2.874008460348417e-05, + "loss": 0.0289, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.14579996466636658, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.12655268609523773, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0246, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.15124735236167908, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0297, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.16273140907287598, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0295, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14014655351638794, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0262, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12657561898231506, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0292, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.13813915848731995, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.0287, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.12874144315719604, + "learning_rate": 2.872678492608539e-05, + "loss": 0.026, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.12086840718984604, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0245, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.15290603041648865, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0297, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.13261260092258453, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0263, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.17371073365211487, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0268, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.17300288379192352, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0264, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.16441872715950012, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0279, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.17883789539337158, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.03, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.14519335329532623, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0266, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.18063998222351074, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0275, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.19818824529647827, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0256, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1695982664823532, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0274, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.16905854642391205, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0277, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.1852719783782959, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.03, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.14601513743400574, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0261, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.11720956116914749, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0279, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.14177089929580688, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0268, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.14933669567108154, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0292, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15183576941490173, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0285, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.1255270391702652, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.025, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.1467735320329666, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0258, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.13636574149131775, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0278, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.12642863392829895, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0254, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.09835807979106903, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0284, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.11784205585718155, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0285, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.14453290402889252, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0301, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.19569514691829681, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0264, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.13860125839710236, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.026, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.16651827096939087, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0283, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.13682647049427032, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0309, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.14632774889469147, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0275, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.1401532143354416, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0261, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.12643243372440338, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0268, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1152961477637291, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0259, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.14027607440948486, + "learning_rate": 2.866951776928e-05, + "loss": 0.0306, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.12379784137010574, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0272, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.15595580637454987, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0288, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.12823393940925598, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0263, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1481468379497528, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0283, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.12768998742103577, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0262, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.15771906077861786, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0267, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.18978171050548553, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0295, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.1386374831199646, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0277, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.1583528220653534, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0273, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.09999477118253708, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0253, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.14432361721992493, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0289, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.19001086056232452, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.0295, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.13931159675121307, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0257, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.1621151715517044, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0269, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.15239235758781433, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0275, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12619581818580627, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0273, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.14698320627212524, + "learning_rate": 2.864043385528954e-05, + "loss": 0.0275, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.14121954143047333, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0271, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.16203583776950836, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0265, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.15562862157821655, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0267, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.16438758373260498, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0285, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.13769859075546265, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0274, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.14908984303474426, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0256, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.14057780802249908, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.0261, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.16056638956069946, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.0267, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.12299710512161255, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0257, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.12815117835998535, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0284, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.14666494727134705, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0303, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.15829558670520782, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0243, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1499258577823639, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.029, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.16584458947181702, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0272, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1765419989824295, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0278, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.17066679894924164, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0281, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13183729350566864, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0255, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.12179457396268845, + "learning_rate": 2.86093128283797e-05, + "loss": 0.028, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.10134928673505783, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0275, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16393089294433594, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0278, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1360260546207428, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0297, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.16158054769039154, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0249, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.12037541717290878, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.0286, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.16677190363407135, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0277, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1442420333623886, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0268, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.1756538599729538, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0282, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.11580783128738403, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0282, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.11671322584152222, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0279, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.15163642168045044, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0297, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.1390339583158493, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0282, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.11784064769744873, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0275, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.15270282328128815, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0282, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.09030280262231827, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0281, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.15108083188533783, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0274, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.1403200924396515, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0265, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.14890284836292267, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0303, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.11579708009958267, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0283, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.14590834081172943, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0268, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.1513289511203766, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0264, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.13494084775447845, + "learning_rate": 2.857082133984191e-05, + "loss": 0.026, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.11738834530115128, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.029, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.1497882753610611, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0271, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.17350555956363678, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0277, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.16718043386936188, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0269, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.12658995389938354, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0267, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15880689024925232, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0256, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.1480095237493515, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0263, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.13779640197753906, + "learning_rate": 2.85567006866004e-05, + "loss": 0.0282, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1088612824678421, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0245, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.17420272529125214, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0267, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.12365500628948212, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.025, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.13261641561985016, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0256, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.10485266894102097, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0253, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14070947468280792, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0284, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.1179763451218605, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0241, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13687312602996826, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0284, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.17160922288894653, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0263, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.17487400770187378, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0275, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.11352817714214325, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1871306300163269, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0277, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.13434743881225586, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0278, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.16361933946609497, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0257, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.19457586109638214, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0277, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.16905367374420166, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0323, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.14845560491085052, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0248, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.12934044003486633, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.026, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.11046823114156723, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0267, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.14174290001392365, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0255, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.15194296836853027, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.0257, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.1532754749059677, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0269, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19864243268966675, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0288, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.16765300929546356, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0271, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.15602844953536987, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0288, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.1571321040391922, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0264, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1625281423330307, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0268, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.12808986008167267, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0307, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.12471310794353485, + "learning_rate": 2.850496147627652e-05, + "loss": 0.027, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.1521507054567337, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0289, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1389119178056717, + "learning_rate": 2.850136142292186e-05, + "loss": 0.024, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1338133066892624, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13214652240276337, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.0274, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.17002162337303162, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0277, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.1433296948671341, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0267, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.19547788798809052, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0287, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.1916888952255249, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0296, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.17188303172588348, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0262, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.14335697889328003, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.0269, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14298376441001892, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0278, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.12892583012580872, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0278, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.16246432065963745, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0255, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.1237327978014946, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0252, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.12249717861413956, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0238, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.16828545928001404, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0262, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.12834671139717102, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0287, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.11583935469388962, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.029, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.10479022562503815, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0244, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.18407097458839417, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0276, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.12898766994476318, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0277, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.13034385442733765, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0241, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.13197654485702515, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0269, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.14327986538410187, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0281, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.15228267014026642, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0269, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.11858481913805008, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0261, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.16075590252876282, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0284, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.1527116447687149, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0283, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.1280883252620697, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0294, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1348433941602707, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0251, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.15039700269699097, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0272, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.16682399809360504, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.0277, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.11187390983104706, + "learning_rate": 2.844503681416359e-05, + "loss": 0.0298, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.1675758808851242, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0267, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.16022244095802307, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0301, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.16192372143268585, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0277, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.13092738389968872, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.12416383624076843, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.028, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.12514744699001312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0252, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.1705007702112198, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.0274, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.14251261949539185, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0264, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.147093266248703, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0279, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.12747326493263245, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0265, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.14222867786884308, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0276, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.12574070692062378, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0262, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.12812504172325134, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0225, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.13585571944713593, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0267, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.12168630957603455, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0236, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.12048719823360443, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0225, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.15165774524211884, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0267, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.15395723283290863, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.026, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.14715734124183655, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0259, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.11370360106229782, + "learning_rate": 2.840817735753433e-05, + "loss": 0.0284, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1264869123697281, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.028, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.13726982474327087, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0268, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1450515240430832, + "learning_rate": 2.840261328154438e-05, + "loss": 0.0251, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.1328781247138977, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.025, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.15474216639995575, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.0292, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.1338219791650772, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0279, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.16368664801120758, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.027, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.18125605583190918, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.0252, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.127846822142601, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0252, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.17743080854415894, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0275, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.16253551840782166, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0287, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.14212867617607117, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0274, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.1380675882101059, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0263, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.12773451209068298, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.0257, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.1556098759174347, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0264, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.16690294444561005, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0296, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.11710768938064575, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0252, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.17441801726818085, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0318, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.10790088772773743, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0251, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.16537238657474518, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0266, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.1414048671722412, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0276, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.1096765398979187, + "learning_rate": 2.836716156259365e-05, + "loss": 0.0258, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.14450953900814056, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0278, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15733887255191803, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0278, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.15253902971744537, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0275, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.1410975307226181, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0249, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1242721900343895, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0251, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.15358713269233704, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0287, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.15856830775737762, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.029, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.1777517944574356, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0254, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.10823221504688263, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0262, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.144780695438385, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0286, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12900282442569733, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0243, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.15370796620845795, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0238, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.15801627933979034, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0259, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.19058606028556824, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0257, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14663439989089966, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0281, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.10609344393014908, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.0261, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.11403398960828781, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0291, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.11631747335195541, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0243, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.1325336992740631, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0266, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.11245301365852356, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.10258981585502625, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0252, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.15847361087799072, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0258, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.15267148613929749, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0246, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.15297843515872955, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0248, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.1767774075269699, + "learning_rate": 2.831995639121568e-05, + "loss": 0.0292, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.12696890532970428, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.025, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.17774799466133118, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0267, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.15603452920913696, + "learning_rate": 2.83142492407012e-05, + "loss": 0.0285, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.12731599807739258, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0272, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.13816066086292267, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0265, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.14216956496238708, + "learning_rate": 2.830853299002291e-05, + "loss": 0.026, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1427476406097412, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0263, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.14667609333992004, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0282, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.11785806715488434, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0276, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.125101238489151, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0255, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.10735481232404709, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0248, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.14256443083286285, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0286, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12700040638446808, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0275, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.11069332808256149, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0283, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.1438131332397461, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.0269, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.1901932805776596, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0274, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.17823027074337006, + "learning_rate": 2.828557706390495e-05, + "loss": 0.029, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.13579529523849487, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0267, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.1431250274181366, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0279, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.11313948035240173, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0274, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1669740527868271, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0288, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.12642791867256165, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0267, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.15632158517837524, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0256, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.1404937356710434, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0263, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.13466882705688477, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0249, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.13014933466911316, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0255, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.18688003718852997, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0267, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.14594615995883942, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0253, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.1708635836839676, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0244, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.16193392872810364, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0265, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.13249778747558594, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0286, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.13875705003738403, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0262, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.19137892127037048, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0246, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.15596376359462738, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0248, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.1577422320842743, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0247, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.14322742819786072, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0263, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.13651713728904724, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0261, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11288725584745407, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0252, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.09636299312114716, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0267, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.11729153245687485, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0275, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.14827747642993927, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0265, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.1446153074502945, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0257, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.1329014152288437, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0247, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.12726181745529175, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0259, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.13508957624435425, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0258, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.14538758993148804, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0264, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.10585593432188034, + "learning_rate": 2.822755216255228e-05, + "loss": 0.0276, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.17435809969902039, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0281, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.1504661589860916, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0259, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.1678706258535385, + "learning_rate": 2.8221699881734e-05, + "loss": 0.029, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.13508610427379608, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0261, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12610375881195068, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0237, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.12272591143846512, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0231, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.1211056336760521, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0252, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.13204888999462128, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0276, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.11982017755508423, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0245, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.17140068113803864, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0236, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.12660211324691772, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0264, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1636001467704773, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0266, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.16939589381217957, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0276, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.15193505585193634, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0291, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16562964022159576, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0283, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.13603557646274567, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.024, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.14359496533870697, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0257, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.14077764749526978, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0247, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.14604462683200836, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0277, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.1240646094083786, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0262, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13144497573375702, + "learning_rate": 2.818639656220465e-05, + "loss": 0.0255, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.1384861171245575, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0227, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.16475403308868408, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0269, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.14310070872306824, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0258, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1498526632785797, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0263, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.13084106147289276, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0268, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.12004188448190689, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0252, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.14093606173992157, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0262, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.11016195267438889, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.0249, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.11101678013801575, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0246, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.1335095763206482, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0274, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1479913741350174, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0237, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.11868918687105179, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0231, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12799321115016937, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.027, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.17011503875255585, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0271, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.15915420651435852, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0274, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.11501206457614899, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0254, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.1457805633544922, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0274, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.1501278430223465, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0264, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.13190113008022308, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0272, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.14392758905887604, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0235, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.1375834047794342, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0241, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.14287824928760529, + "learning_rate": 2.814280752477756e-05, + "loss": 0.0269, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12234462797641754, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0259, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.1336476355791092, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0242, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.11315035820007324, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0245, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.14132073521614075, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0293, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.12053677439689636, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.0266, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.10896297544240952, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0268, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.11100424081087112, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.026, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.11256317049264908, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0237, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.14322060346603394, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0255, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.1655588448047638, + "learning_rate": 2.812283453196085e-05, + "loss": 0.026, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.14714404940605164, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0261, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.12317050993442535, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0247, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12872155010700226, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0258, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.1666097790002823, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0261, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.13425199687480927, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.16131597757339478, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0295, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.13762447237968445, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0242, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.12883082032203674, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0274, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.15029405057430267, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0269, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.13236507773399353, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0303, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.12054237723350525, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0218, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.1471109539270401, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0242, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.11926044523715973, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0275, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.13649292290210724, + "learning_rate": 2.809470494867301e-05, + "loss": 0.0231, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.15434928238391876, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0283, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.11536315083503723, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0271, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.12312797456979752, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0249, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.12750110030174255, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0235, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.1568569540977478, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0277, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.11667264997959137, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0258, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.14598897099494934, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0266, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.21723008155822754, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0258, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.14233075082302094, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0244, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.14836373925209045, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0265, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.146131694316864, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0258, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.16575932502746582, + "learning_rate": 2.807043872632495e-05, + "loss": 0.025, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.15142853558063507, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0245, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.11762700974941254, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0253, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.09714391082525253, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0233, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.1586424708366394, + "learning_rate": 2.806231820878714e-05, + "loss": 0.023, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.14317330718040466, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.0271, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.10633813589811325, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.0244, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.19832657277584076, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0249, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.14428019523620605, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0283, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.18295790255069733, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.0257, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.16204482316970825, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0285, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.14742101728916168, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0239, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1435142159461975, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0253, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.2000540792942047, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.16311310231685638, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0253, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.1238052025437355, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0249, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.16693037748336792, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0269, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.12382039427757263, + "learning_rate": 2.803581696238626e-05, + "loss": 0.025, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.13866499066352844, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0259, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.12111294269561768, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0258, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.13717366755008698, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0266, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.1119573786854744, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0264, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.1390918791294098, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.12881262600421906, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0251, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.1354016363620758, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0256, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.1285468488931656, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0315, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.17634150385856628, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0254, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.15077067911624908, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.0268, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16464342176914215, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0265, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.3138365149497986, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0255, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.15954458713531494, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0272, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15878041088581085, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0267, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.17743942141532898, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0273, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.16661205887794495, + "learning_rate": 2.800297040373295e-05, + "loss": 0.0268, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.13536596298217773, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.0241, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.13258172571659088, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0266, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.1524200588464737, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0276, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.12779775261878967, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0245, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1008654311299324, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0222, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.14413924515247345, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0297, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.14558148384094238, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0252, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.15591812133789062, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0248, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.12246204912662506, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0276, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.13106316328048706, + "learning_rate": 2.7982312890369e-05, + "loss": 0.0265, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.12663032114505768, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0234, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.1167084351181984, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15868185460567474, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0251, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13669976592063904, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0291, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.15100552141666412, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0237, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14283764362335205, + "learning_rate": 2.796987104810136e-05, + "loss": 0.024, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.13055892288684845, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0255, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.16733981668949127, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0244, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.1716628074645996, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0257, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.11838851124048233, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0288, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.14557431638240814, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0238, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.14981132745742798, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0254, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.11149631440639496, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0256, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.20500904321670532, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0255, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.12128293514251709, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0254, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.12934274971485138, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0254, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.1355874240398407, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0267, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13485173881053925, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0259, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.1481594294309616, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0251, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.15145331621170044, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0243, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.16746139526367188, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0284, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.10760390013456345, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0268, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.1460299789905548, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0264, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.15936219692230225, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0232, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.1373136043548584, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.0255, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.14508306980133057, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0243, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.153093621134758, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0262, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.15373852849006653, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0238, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.14033931493759155, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0267, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.1424216777086258, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0252, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.11820772290229797, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0253, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.13179944455623627, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0243, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.18073157966136932, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0256, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.12908786535263062, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0254, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.18455258011817932, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0263, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.14844198524951935, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0248, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.14291873574256897, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0268, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.105900838971138, + "learning_rate": 2.79029165248073e-05, + "loss": 0.027, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.14083242416381836, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0244, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.14579042792320251, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0239, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.13705924153327942, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0254, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.17981873452663422, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0225, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.13812783360481262, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0278, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12511365115642548, + "learning_rate": 2.789025068651819e-05, + "loss": 0.027, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.13009224832057953, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0245, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.145152747631073, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0281, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.1607820838689804, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0254, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.09400778263807297, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0258, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.1608203500509262, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0263, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.14968201518058777, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0247, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.1333402544260025, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0272, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.17871612310409546, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0256, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.10688412934541702, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0264, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.154439315199852, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0259, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.1270124316215515, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0279, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.15214042365550995, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0259, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.14094167947769165, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0261, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.11419129371643066, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.025, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.1447947919368744, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0258, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.12327348440885544, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0246, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.11033894866704941, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0259, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.128106951713562, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0247, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.13206088542938232, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0254, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.13664662837982178, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0274, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1349778026342392, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0262, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.10689893364906311, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0262, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.16241611540317535, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0269, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14955222606658936, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0251, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.13519948720932007, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0231, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.15076202154159546, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0272, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.15367662906646729, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0255, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.15819156169891357, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0242, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.12844862043857574, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0259, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.13392682373523712, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0245, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.15134917199611664, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.0287, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.14139734208583832, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0254, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.1562367081642151, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0246, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.1298898160457611, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0252, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13048380613327026, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0252, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.12535928189754486, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0271, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.19301816821098328, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0245, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.12262707948684692, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0262, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1394580751657486, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0272, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.21376213431358337, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0243, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.15442071855068207, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0285, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.13063938915729523, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0277, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.15023048222064972, + "learning_rate": 2.779844963973707e-05, + "loss": 0.025, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.17523249983787537, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0239, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.1312434822320938, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0234, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13352124392986298, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0249, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.12799884378910065, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0257, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.12704412639141083, + "learning_rate": 2.778765821684581e-05, + "loss": 0.025, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.11119242757558823, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0248, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.1564178764820099, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0282, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.13500694930553436, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0279, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.1542324423789978, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0258, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13888850808143616, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0281, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.14011257886886597, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.14887000620365143, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0254, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.15670792758464813, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0289, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.1288663148880005, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0242, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.13191169500350952, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.027, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.1320677250623703, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0239, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.13418279588222504, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0234, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.11019188910722733, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0232, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.13963419198989868, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0235, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.1163327619433403, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.028, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.13508521020412445, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0275, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.11098913848400116, + "learning_rate": 2.775078589916535e-05, + "loss": 0.025, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.16348809003829956, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0259, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.12986604869365692, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0246, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1302310824394226, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0259, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.11797303706407547, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.0252, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.1427793651819229, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0233, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.13411074876785278, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.0257, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.1528121531009674, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.11762527376413345, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0238, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.13639862835407257, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0257, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.1356336623430252, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0269, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.16879825294017792, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.13352762162685394, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0236, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.20271515846252441, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0296, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.11791368573904037, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0241, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14179861545562744, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.0256, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12654612958431244, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0243, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12589605152606964, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0222, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.12714968621730804, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0278, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.13410630822181702, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0248, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.12528881430625916, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0258, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.11806205660104752, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0308, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.14317870140075684, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0244, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14099767804145813, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0228, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.15186408162117004, + "learning_rate": 2.769825463131276e-05, + "loss": 0.0252, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.18532416224479675, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0262, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.1445159614086151, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0265, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14576950669288635, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0253, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1777651309967041, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0229, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.16440069675445557, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0234, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.12858232855796814, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0259, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.14329349994659424, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0254, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.1265982836484909, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0235, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13783425092697144, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0228, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.1327633559703827, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0248, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14867408573627472, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0257, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.1556752771139145, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0239, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.16407394409179688, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0261, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.14142248034477234, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0265, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.11142456531524658, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0221, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.127807155251503, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0265, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.14716863632202148, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0241, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.1455392986536026, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.024, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1665087193250656, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0253, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.16937066614627838, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0292, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.13188236951828003, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0254, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.1010027602314949, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.0236, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.15787169337272644, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.0279, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.169437438249588, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0246, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.22705122828483582, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0227, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.17403557896614075, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.027, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.1287415474653244, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0253, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.13813741505146027, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0238, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.11655540764331818, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.13267700374126434, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.025, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.10447681695222855, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0239, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.11865647882223129, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.025, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.13245069980621338, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0256, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.11377252638339996, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0266, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.11403302848339081, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0252, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.1518944650888443, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0286, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.13505733013153076, + "learning_rate": 2.761618148370765e-05, + "loss": 0.0253, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.12230793386697769, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0247, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.1600245237350464, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0267, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.13100600242614746, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.11071992665529251, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0254, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.16208025813102722, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.0287, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.1652526706457138, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.027, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14042513072490692, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.1674395203590393, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0257, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.138246089220047, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0272, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13758982717990875, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0227, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.15503156185150146, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0259, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.15050673484802246, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0279, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.13952329754829407, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.024, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.14903564751148224, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0242, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.09984514862298965, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.09952808916568756, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0233, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.14514309167861938, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0255, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.12824438512325287, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0232, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.15507018566131592, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0251, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.12279994785785675, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0225, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.13601353764533997, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0249, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.20022229850292206, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.0263, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.1407146006822586, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0226, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.1350381076335907, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0248, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.13220156729221344, + "learning_rate": 2.755998359318217e-05, + "loss": 0.0251, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.11746657639741898, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0256, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.10933379083871841, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0235, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.1263892501592636, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0244, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.14062181115150452, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.0255, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.1430615335702896, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0254, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.19069550931453705, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0254, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.16568611562252045, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0252, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.15911205112934113, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.0265, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.17054520547389984, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0265, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.15547682344913483, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0237, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.12778308987617493, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0242, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.1590787172317505, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0232, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13945519924163818, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.023, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.14507925510406494, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0247, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14541085064411163, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0245, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.17666012048721313, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0222, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.16692468523979187, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0218, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.1295796036720276, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0242, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.12527599930763245, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0273, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.13334719836711884, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0253, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.1175302267074585, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0251, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.14461147785186768, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0261, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.15658612549304962, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0236, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.15802229940891266, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0236, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.14738211035728455, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0252, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.1639637053012848, + "learning_rate": 2.750090542643057e-05, + "loss": 0.027, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1573086827993393, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0264, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.13114139437675476, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0238, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.1112227514386177, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.027, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.12209689617156982, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.13741743564605713, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0239, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14444704353809357, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0263, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.13822096586227417, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0246, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.14305464923381805, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0237, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13539855182170868, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0264, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.13791607320308685, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0234, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.205885648727417, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0234, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.1034194752573967, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0214, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.15530072152614594, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0258, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.1535874903202057, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0226, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.19778263568878174, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0238, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.175798699259758, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0255, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.1803887039422989, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0238, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.1906212419271469, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0253, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.13742655515670776, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.026, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.12772122025489807, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0261, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13805702328681946, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0259, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.14081542193889618, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0241, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.10023088753223419, + "learning_rate": 2.744810891043728e-05, + "loss": 0.0246, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.10111884027719498, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0215, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.11498598754405975, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0217, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10912947356700897, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.0257, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.13553178310394287, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0266, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.18210133910179138, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0234, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.11901862919330597, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0248, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.14622537791728973, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0234, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.1641179323196411, + "learning_rate": 2.742962757296782e-05, + "loss": 0.026, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.14369463920593262, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.0236, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.14231407642364502, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0213, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.10929716378450394, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0216, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.13811182975769043, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0241, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.13219693303108215, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0219, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.13322590291500092, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0259, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.14716853201389313, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0255, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.11750922352075577, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0272, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.1430707424879074, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1722290813922882, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.025, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.15502606332302094, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.0231, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.09124620258808136, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0246, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.16296057403087616, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0277, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.10881830751895905, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0222, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.11321897804737091, + "learning_rate": 2.739481230505444e-05, + "loss": 0.026, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.1362120658159256, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.03, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.14329490065574646, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0252, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.4512016475200653, + "learning_rate": 2.738782381719405e-05, + "loss": 0.0238, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12559202313423157, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0263, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.12616541981697083, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0248, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.1516106277704239, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0234, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.12785211205482483, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0264, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16447393596172333, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.026, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.1648014783859253, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0264, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.14414788782596588, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0249, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.10982059687376022, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0254, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.12677313387393951, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0263, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.12355965375900269, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0223, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.15190085768699646, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0246, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.13910633325576782, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0231, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.11962535977363586, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0259, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.1566285789012909, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0256, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.12282316386699677, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0253, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.16674686968326569, + "learning_rate": 2.735040900116229e-05, + "loss": 0.025, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.14292193949222565, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0277, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.1504344791173935, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0268, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.16317704319953918, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0229, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.20625433325767517, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0242, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.1895458698272705, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.024, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.1987384408712387, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.0258, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.18668580055236816, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.1807073950767517, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0246, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.11927719414234161, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.028, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.11955022811889648, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0259, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.15261328220367432, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0268, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.15620580315589905, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0254, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.12000948935747147, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0234, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.1188279539346695, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.0248, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.13279424607753754, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0241, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.13379216194152832, + "learning_rate": 2.731275407491117e-05, + "loss": 0.026, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.12054166197776794, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0215, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.15941590070724487, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0279, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.18678002059459686, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0263, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.21807490289211273, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.0239, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.12243734300136566, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.028, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.14628900587558746, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0248, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.17471152544021606, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0257, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.11189062893390656, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0242, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.16700562834739685, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0252, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.1213584691286087, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0251, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.14496786892414093, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0258, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.17899028956890106, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0248, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13594108819961548, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0246, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.1365566849708557, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.021, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.10815171152353287, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0227, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.11390877515077591, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0209, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.1541927456855774, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0213, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.1342182159423828, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0251, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.1300208568572998, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0249, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14917100965976715, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0274, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.14349427819252014, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.0269, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10514543205499649, + "learning_rate": 2.72605878418951e-05, + "loss": 0.024, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.12946246564388275, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0234, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.1120578870177269, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.0246, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.14252731204032898, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.0287, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.0984109416604042, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0234, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.13058412075042725, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0235, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.1468561738729477, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0263, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.1297929584980011, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0224, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.15071524679660797, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0254, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.15211531519889832, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0239, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.13758648931980133, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0254, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.13972750306129456, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0239, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1762881577014923, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0235, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.13277511298656464, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.025, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.11352386325597763, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0225, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.1340995877981186, + "learning_rate": 2.722476145603374e-05, + "loss": 0.0232, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.16372360289096832, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0224, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.130090594291687, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0231, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.10481512546539307, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0229, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.1453929841518402, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0255, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.133439838886261, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0248, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.1380738615989685, + "learning_rate": 2.721037237880308e-05, + "loss": 0.023, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.1386265754699707, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0264, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.14298051595687866, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.0247, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.11756420135498047, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0222, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1197751984000206, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0216, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.2194000482559204, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0243, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.14157775044441223, + "learning_rate": 2.719594991888059e-05, + "loss": 0.023, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.14389759302139282, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0228, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.12442908436059952, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0267, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.1188371554017067, + "learning_rate": 2.718872618272836e-05, + "loss": 0.027, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11137005686759949, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0237, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.10907725989818573, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0251, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.150213360786438, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0229, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.2098330855369568, + "learning_rate": 2.717908157623383e-05, + "loss": 0.0245, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.12558938562870026, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0257, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.13659170269966125, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0247, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.12037219852209091, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0228, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.13085514307022095, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0252, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.13916544616222382, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0251, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.13098886609077454, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.021, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.12777045369148254, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0258, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.10204357653856277, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0249, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.11189272999763489, + "learning_rate": 2.715732712163776e-05, + "loss": 0.0232, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.1412421613931656, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0255, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.09470678865909576, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0217, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.09736761450767517, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0233, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12988215684890747, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.1080017015337944, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0263, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.12020932137966156, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0238, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.1058862954378128, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0247, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11686211824417114, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.024, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.11078432947397232, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0226, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.13394472002983093, + "learning_rate": 2.713306780076029e-05, + "loss": 0.0237, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.11299733072519302, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.025, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.1196199283003807, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0251, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.16477037966251373, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0241, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.1175335943698883, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.11024156212806702, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0237, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.12922869622707367, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0278, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.1429518610239029, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0254, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.14210145175457, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0225, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.12270870059728622, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0257, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.12436466664075851, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0258, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.10704488307237625, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0224, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.15890415012836456, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0252, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.12372975796461105, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0225, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.11652565747499466, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.028, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.1483478546142578, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0255, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.13188686966896057, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0238, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.18725869059562683, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0244, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.15949182212352753, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0225, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.14752954244613647, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0229, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.14210951328277588, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0233, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.1388787031173706, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.0253, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.11010867357254028, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0227, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.13336122035980225, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0227, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.10118173807859421, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.0265, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.1434611827135086, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0263, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.14133326709270477, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.026, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.15359432995319366, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0236, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.12927602231502533, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.022, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.10330940783023834, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0236, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.1264553815126419, + "learning_rate": 2.705973772304842e-05, + "loss": 0.025, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.15007789433002472, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0267, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.11090201884508133, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0257, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.12769529223442078, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0252, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.0977141410112381, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0244, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.14598220586776733, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0254, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.11231301724910736, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0235, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.09610529243946075, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0262, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.1549275815486908, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0234, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.11284032464027405, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.0217, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.14465086162090302, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.0235, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.13792403042316437, + "learning_rate": 2.703264323632037e-05, + "loss": 0.024, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.1438492238521576, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0239, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.12613703310489655, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0244, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.14072397351264954, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0225, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.1420842856168747, + "learning_rate": 2.702276326859174e-05, + "loss": 0.026, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.17447218298912048, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0243, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.1708984673023224, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0259, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.1555069535970688, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.0252, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.12473946064710617, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0227, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.15750949084758759, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0248, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.14009647071361542, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0267, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.1416911780834198, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0229, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.11944608390331268, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0235, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.126204252243042, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0268, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.1186814159154892, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0242, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.12699876725673676, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0239, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.10452742129564285, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0221, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.1422647386789322, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0247, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.11601094901561737, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0232, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12339281290769577, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0223, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1883407086133957, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0243, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.1730756312608719, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0294, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.10429747402667999, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0194, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.1358909010887146, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0217, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.14417894184589386, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0265, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.12299799919128418, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0236, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.1361788511276245, + "learning_rate": 2.696816263672279e-05, + "loss": 0.024, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.14492252469062805, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0241, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.1209225058555603, + "learning_rate": 2.696317709902802e-05, + "loss": 0.0239, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.11048539727926254, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.12423776835203171, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0243, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.1254356950521469, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.12912537157535553, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0254, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.10864657163619995, + "learning_rate": 2.695069735893089e-05, + "loss": 0.022, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.15519478917121887, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0253, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.15276901423931122, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0231, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.1401076465845108, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0271, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.14237190783023834, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.024, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1186559721827507, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0238, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.1254940629005432, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0233, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.1337885558605194, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0255, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12958411872386932, + "learning_rate": 2.69306825906724e-05, + "loss": 0.025, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.11610005795955658, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0238, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.13515442609786987, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0239, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1492868959903717, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0224, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.1293300986289978, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.021, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12354710698127747, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.13226614892482758, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.0247, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.1368253529071808, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0255, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.13044875860214233, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0228, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14107753336429596, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0221, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.13521252572536469, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0236, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.13718804717063904, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0217, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.15192179381847382, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.0247, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.13718710839748383, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0244, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11781664192676544, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0239, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.14431379735469818, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0248, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.15059512853622437, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0231, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.12076479196548462, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0242, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.19006308913230896, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0238, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.140592560172081, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0233, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.1528412401676178, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0229, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.16418038308620453, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0242, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.12185576558113098, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0254, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11594245582818985, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0248, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1216709092259407, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0234, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.12024594098329544, + "learning_rate": 2.686776313743295e-05, + "loss": 0.024, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.1321716010570526, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.0255, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.13785192370414734, + "learning_rate": 2.686270520700719e-05, + "loss": 0.025, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.12321551889181137, + "learning_rate": 2.6860174890357e-05, + "loss": 0.0233, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14606279134750366, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0249, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.17512187361717224, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0272, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.14974310994148254, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.0243, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.11758870631456375, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0241, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.11157143115997314, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0232, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.10899091511964798, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0231, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.1136092096567154, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0224, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.10306857526302338, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0211, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13241226971149445, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0222, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.11886361986398697, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0244, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.11962216347455978, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0237, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.14516247808933258, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.14592237770557404, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0253, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1278424859046936, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0233, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.12451537698507309, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0218, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.11746378242969513, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0266, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.14618739485740662, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0226, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.1512313336133957, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0233, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.14035750925540924, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0264, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.13232699036598206, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.023, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.14616237580776215, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.0233, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.12882372736930847, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0265, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.13778620958328247, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.024, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.15017054975032806, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0239, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.14781621098518372, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0215, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1670294851064682, + "learning_rate": 2.679407107451536e-05, + "loss": 0.022, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.1112075075507164, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0232, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.12219138443470001, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0202, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12922540307044983, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.11967732757329941, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0226, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.1377563774585724, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0224, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.14771930873394012, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0234, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.14800170063972473, + "learning_rate": 2.677617030053916e-05, + "loss": 0.0251, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.1233784481883049, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0232, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.16905057430267334, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0214, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.13751909136772156, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0248, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.13289029896259308, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0216, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.10788674652576447, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0216, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.13367260992527008, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1270209699869156, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0239, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.140412375330925, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0218, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.17788565158843994, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0253, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.13695330917835236, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.023, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.1250872015953064, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0232, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.11347977817058563, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0243, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.15546919405460358, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0233, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.13752058148384094, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.10431966185569763, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0215, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.17878347635269165, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0262, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.13218404352664948, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0241, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.12244930118322372, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.0247, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.11582386493682861, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0219, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.1160157322883606, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0203, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.10731764137744904, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0234, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.11152671277523041, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0218, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.12760375440120697, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0218, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.13706916570663452, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0219, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.14939993619918823, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0208, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.1169784739613533, + "learning_rate": 2.67092985767871e-05, + "loss": 0.0217, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.1569128930568695, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12233320623636246, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0251, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13152818381786346, + "learning_rate": 2.670154386126402e-05, + "loss": 0.0231, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.1529632955789566, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0246, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.11001774668693542, + "learning_rate": 2.669636960731296e-05, + "loss": 0.0223, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.1473267376422882, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.025, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.12853027880191803, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0246, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.15698745846748352, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0209, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12175270169973373, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0229, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.15634675323963165, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.0224, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.16060872375965118, + "learning_rate": 2.668082553346281e-05, + "loss": 0.0254, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.13773131370544434, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0224, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.14258436858654022, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0216, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.13032476603984833, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0215, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.17300327122211456, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0221, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.13187018036842346, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0226, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.14065369963645935, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0232, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.11484711617231369, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0263, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1380661576986313, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0209, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.17800134420394897, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0244, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.14836716651916504, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0277, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14563506841659546, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.023, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.11463280767202377, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.026, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.13092686235904694, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0222, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.12265385687351227, + "learning_rate": 2.664443191002912e-05, + "loss": 0.025, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.14121955633163452, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0242, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.12938837707042694, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0222, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.1922816038131714, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.0228, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.12596914172172546, + "learning_rate": 2.663400187275054e-05, + "loss": 0.022, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.14095240831375122, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0244, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.1080891340970993, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0228, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.13916714489459991, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.0239, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.1489475965499878, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0251, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.11781706660985947, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0207, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.11544635146856308, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0232, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.14827974140644073, + "learning_rate": 2.661571529941175e-05, + "loss": 0.0236, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.12821446359157562, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.0186, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.1303764134645462, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0234, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.1049366369843483, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0232, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.15925322473049164, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0275, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.13151033222675323, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0213, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.09425143152475357, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0215, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.15108847618103027, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0228, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.1374080330133438, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0231, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.16061817109584808, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0253, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.1707434207201004, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0259, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.12051897495985031, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0217, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.09191743284463882, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0224, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.12547235190868378, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0245, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14491257071495056, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0228, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.11896242201328278, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0239, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.14253611862659454, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0233, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.12803201377391815, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.0226, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.12893305718898773, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0231, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.16127154231071472, + "learning_rate": 2.656586262343925e-05, + "loss": 0.0234, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.13307608664035797, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0258, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.11211924254894257, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.0219, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.1672789305448532, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0248, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.12878142297267914, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0231, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.11429315060377121, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0233, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.13469499349594116, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0254, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.12753085792064667, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0258, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.11875704675912857, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0234, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.13718342781066895, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0223, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.10762278735637665, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.021, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.11651597917079926, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0222, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.12953802943229675, + "learning_rate": 2.653421326308085e-05, + "loss": 0.0235, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.14820824563503265, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.023, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.16258545219898224, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0266, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.14097236096858978, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0224, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.1039188802242279, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0216, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15014079213142395, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0246, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12850438058376312, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.023, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.12245986610651016, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0209, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.15426407754421234, + "learning_rate": 2.651304359235998e-05, + "loss": 0.025, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.13411583006381989, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0229, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.13705122470855713, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0254, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.13120216131210327, + "learning_rate": 2.650509053370553e-05, + "loss": 0.022, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14599482715129852, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0249, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.11628340184688568, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0215, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.13643144071102142, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0228, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.12846967577934265, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0214, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.13181492686271667, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0245, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1277213841676712, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0244, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12288910150527954, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0226, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.14789454638957977, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0215, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.11457780748605728, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0248, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.12511487305164337, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0206, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.12255935370922089, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0245, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.13214053213596344, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0238, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.1601077914237976, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.027, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.1244364008307457, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0212, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.11869437992572784, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0239, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.14327670633792877, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0227, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.12729141116142273, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0248, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.10953658074140549, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0205, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.20326738059520721, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0225, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.15614672005176544, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0237, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.11863692849874496, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0208, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.10683806240558624, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0224, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.1529647260904312, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.024, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.18428578972816467, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0232, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.19935141503810883, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.0223, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.14675194025039673, + "learning_rate": 2.643583521046136e-05, + "loss": 0.0216, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.21206484735012054, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0243, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.14521193504333496, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0214, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.11565426737070084, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0236, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.1087750792503357, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0215, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.11942578107118607, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0229, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.138235405087471, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0219, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.1079302579164505, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0222, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.15928085148334503, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0225, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.10920153558254242, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.024, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.12247785925865173, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0205, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.10454121232032776, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0238, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.1726839691400528, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0211, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1617811918258667, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0225, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.15452679991722107, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0214, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1573309749364853, + "learning_rate": 2.639561293166059e-05, + "loss": 0.022, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.13596323132514954, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0234, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.09188699722290039, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0219, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.12158972024917603, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0202, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.16615170240402222, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.0221, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.1163380891084671, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0255, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.17704762518405914, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0243, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.1591942012310028, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.021, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.14697043597698212, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0241, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1234351247549057, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0261, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.1379440575838089, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0266, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1025586649775505, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.022, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.11894206702709198, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0226, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1431341916322708, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0241, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.10291697829961777, + "learning_rate": 2.635789644394928e-05, + "loss": 0.022, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.1622016727924347, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0229, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.11867184937000275, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0262, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.18177814781665802, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0237, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.15159954130649567, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0218, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.11481288820505142, + "learning_rate": 2.634438526911474e-05, + "loss": 0.024, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.10989677160978317, + "learning_rate": 2.634168044873305e-05, + "loss": 0.024, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.12370797991752625, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0235, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1390318125486374, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0231, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.14030151069164276, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.0207, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11659140884876251, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.021, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.16727864742279053, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0242, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.11524021625518799, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0265, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.10776443779468536, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0224, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.16366012394428253, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0264, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.1102370023727417, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0245, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11543556302785873, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0216, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.11101235449314117, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0208, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.1257167011499405, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.024, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.1063276082277298, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0241, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.14303457736968994, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0242, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.12862786650657654, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0244, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15487778186798096, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.14244592189788818, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0211, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1066693440079689, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0199, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11721482872962952, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0214, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.13803230226039886, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.0227, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.11225558817386627, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.028, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.11660541594028473, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0239, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12551093101501465, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0221, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.1305139809846878, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.1281510442495346, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0235, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.14735114574432373, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0213, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.10699455440044403, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0231, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.16579580307006836, + "learning_rate": 2.626559653214893e-05, + "loss": 0.0213, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.09389153122901917, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0218, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.17565062642097473, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0245, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.1368194818496704, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0214, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.10599412024021149, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0221, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.12645669281482697, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0233, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.12325828522443771, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.0228, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.11353788524866104, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.10889376699924469, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0227, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.12935040891170502, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0214, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.1422247439622879, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0219, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.16723784804344177, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0239, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.09775881469249725, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.0229, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.13455381989479065, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0253, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.13700921833515167, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0239, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12252738326787949, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0231, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.15983854234218597, + "learning_rate": 2.622181856488204e-05, + "loss": 0.0223, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.18716242909431458, + "learning_rate": 2.621907518918971e-05, + "loss": 0.025, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.16380536556243896, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0229, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.1524866819381714, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0221, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.12281060218811035, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0213, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.09793791174888611, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0249, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.12271476536989212, + "learning_rate": 2.620534553467634e-05, + "loss": 0.0221, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.12616175413131714, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0236, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.10274055600166321, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0205, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.16873067617416382, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.027, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.10157257318496704, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0236, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.11863094568252563, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0225, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.09981630742549896, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0233, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11045375466346741, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0226, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.13111108541488647, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0232, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.12417030334472656, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0226, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.12029916048049927, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0217, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.10766672343015671, + "learning_rate": 2.617506544598029e-05, + "loss": 0.025, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.12039721757173538, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0228, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.11044551432132721, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0223, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.12463348358869553, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.0232, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.14382848143577576, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0238, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.13677646219730377, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0239, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.11536712944507599, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0224, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.12461430579423904, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0253, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.16901540756225586, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0242, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.12255802005529404, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0212, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.14185196161270142, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.0218, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.1292722374200821, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0249, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12001507729291916, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0222, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.12286248803138733, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0211, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.12487491965293884, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0231, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13417202234268188, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0257, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.11690660566091537, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0219, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15474916994571686, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0223, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.1720103919506073, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0239, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.16786889731884003, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0216, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.2279045283794403, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.0254, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13872499763965607, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0238, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.1881643831729889, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0211, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.1372661143541336, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0237, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.10514581203460693, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0217, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.11103228479623795, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0248, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1349911093711853, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0237, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.12774604558944702, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0245, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.15125125646591187, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0233, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.1190832257270813, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0248, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.1088283434510231, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0187, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.12167634069919586, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.023, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.15135347843170166, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0226, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.12857669591903687, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.0195, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.13588359951972961, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0234, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.14902308583259583, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0235, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.15095068514347076, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0209, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.1116727739572525, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0255, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11652971059083939, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0211, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.11875079572200775, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0223, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.11100098490715027, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.022, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.13648055493831635, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0234, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.12870723009109497, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0205, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.10456091910600662, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0232, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.13770641386508942, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0212, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.1526155173778534, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0244, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12664297223091125, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0241, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.12000498175621033, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0239, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.11930232495069504, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.0207, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.11384909600019455, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0217, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.1746685951948166, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.023, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.15868088603019714, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0233, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.15131928026676178, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0194, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.10279352962970734, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0215, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.13810355961322784, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0202, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.13024795055389404, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0217, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1374605894088745, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0215, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.14356477558612823, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0206, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.11046172678470612, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0236, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.10146678239107132, + "learning_rate": 2.601090850395125e-05, + "loss": 0.021, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13247323036193848, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0217, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.13672757148742676, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0201, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.08651759475469589, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.0196, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.14821568131446838, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0223, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11985234171152115, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0208, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11231720447540283, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0219, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.12338588386774063, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0202, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.15079368650913239, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0229, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.1348753422498703, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0209, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1333470344543457, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0207, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.10075798630714417, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0232, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1217522993683815, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0247, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.12041056901216507, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.022, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.10997799783945084, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0226, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.119750015437603, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0239, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.15246307849884033, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0239, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.1431925892829895, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0227, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.1181643158197403, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0227, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.13645906746387482, + "learning_rate": 2.595742267991401e-05, + "loss": 0.0213, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.12661530077457428, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0233, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.1344967484474182, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.0221, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.13433857262134552, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0238, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.10517163574695587, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0237, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.10689525306224823, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.022, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.13307900726795197, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0209, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.14696283638477325, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0261, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.12473653256893158, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0213, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.106630340218544, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0221, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.12092334777116776, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0232, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.11033511906862259, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0193, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11916322261095047, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0212, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.1204444020986557, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0228, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.1373489499092102, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0225, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.13124211132526398, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0208, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.18858365714550018, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0258, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.1558879315853119, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0228, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.1481737494468689, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0242, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.10424906760454178, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.021, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.15494640171527863, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0238, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.1801268309354782, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0238, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.14176428318023682, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0228, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.15850430727005005, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0235, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.14727924764156342, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0232, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.14907293021678925, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0242, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.11011895537376404, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0231, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.13541805744171143, + "learning_rate": 2.588090010443142e-05, + "loss": 0.0224, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.13078846037387848, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0228, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.14508767426013947, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0207, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.12445177882909775, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0227, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.14860141277313232, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0244, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.13311992585659027, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0207, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.1157015860080719, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0209, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.13310302793979645, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.0234, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.14026974141597748, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0246, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.14788632094860077, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0213, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.1551811844110489, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0241, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.10706475377082825, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.021, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.12536053359508514, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0228, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.1260751187801361, + "learning_rate": 2.584384079351628e-05, + "loss": 0.022, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.1082935482263565, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0207, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.13922622799873352, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0251, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1240227147936821, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0223, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.14512665569782257, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0218, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12859801948070526, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.0208, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.13418391346931458, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.021, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.10546918958425522, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.022, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.10668392479419708, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10407276451587677, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.02, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.13980497419834137, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0235, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.16233955323696136, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0224, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.16743852198123932, + "learning_rate": 2.580950866038058e-05, + "loss": 0.022, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.09152300655841827, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0206, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.14319349825382233, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0205, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.1045396700501442, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0219, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.10865644365549088, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0201, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.12166012078523636, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0217, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.12505589425563812, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0224, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.1353408247232437, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0242, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.1379590630531311, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0221, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12337373197078705, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0235, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.12092723697423935, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0233, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.09944337606430054, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0215, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.15161453187465668, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0193, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.0963459238409996, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0225, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.15534570813179016, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0247, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.142191082239151, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0228, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.1454230695962906, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0226, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.12661893665790558, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0211, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.11373333632946014, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0213, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12031935155391693, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.0208, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.13401862978935242, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0224, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.11285468935966492, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0231, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.12453637272119522, + "learning_rate": 2.574625966768285e-05, + "loss": 0.0211, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.13041555881500244, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1391122043132782, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0224, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.1345762014389038, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0244, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.12814867496490479, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0219, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.11058632284402847, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0191, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.15668845176696777, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.021, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.1509203314781189, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.0254, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.13209059834480286, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0201, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.1357070952653885, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0226, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.0896834284067154, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0207, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.1570383608341217, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0234, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.16916486620903015, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0254, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.1385425180196762, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.0218, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.11236828565597534, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0214, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11224918812513351, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0233, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.1305687576532364, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0219, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.15051157772541046, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0226, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.13418416678905487, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0244, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.14324583113193512, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0235, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.10810746997594833, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0197, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.14513953030109406, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0215, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1187531128525734, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0193, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11865831911563873, + "learning_rate": 2.567971342427169e-05, + "loss": 0.021, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.12040061503648758, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0229, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.10597001016139984, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0209, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.1219123974442482, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0242, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.11732123047113419, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.0203, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.15642790496349335, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0226, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.1893257051706314, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0247, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11829031258821487, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0239, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.14971555769443512, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0253, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.12424097210168839, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0251, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1071329414844513, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0271, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.11146610230207443, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0214, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.10925749689340591, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.0212, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.14797845482826233, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0221, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.10348086059093475, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0218, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.14607182145118713, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0217, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.11097954958677292, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0259, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.11300431936979294, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0197, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.11342521011829376, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0202, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.12889035046100616, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0217, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.10449834913015366, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0216, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1293669193983078, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0227, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.10667148977518082, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.021, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.0839434340596199, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0236, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.12906265258789062, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0223, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.1100582480430603, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0243, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.10644520074129105, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.0215, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.11693669110536575, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.022, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.13665053248405457, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0227, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12752027809619904, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0241, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.11534811556339264, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0234, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.1467696875333786, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0245, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10596702992916107, + "learning_rate": 2.558641420481684e-05, + "loss": 0.02, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.14685864746570587, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.022, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.13554386794567108, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.026, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.12572740018367767, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0232, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.1583555042743683, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0209, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.12913484871387482, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0226, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.11418721824884415, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0225, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.12722279131412506, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.0211, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.1382816582918167, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.0243, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.11612103134393692, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0212, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.1537163406610489, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.021, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.10462892800569534, + "learning_rate": 2.555415209239832e-05, + "loss": 0.0214, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.12759479880332947, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0232, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.1290607452392578, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0222, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12458004057407379, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0213, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.1113182008266449, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0223, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.15024979412555695, + "learning_rate": 2.553945541728904e-05, + "loss": 0.022, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.12657828629016876, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.025, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.13719463348388672, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.024, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.13465221226215363, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0254, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.08891251683235168, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0215, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.08946283161640167, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0224, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.13241387903690338, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0229, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.13068708777427673, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0215, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.12115749716758728, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.025, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.12863430380821228, + "learning_rate": 2.551295099554365e-05, + "loss": 0.0202, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.1332341581583023, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0204, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.102240651845932, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0199, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.16963273286819458, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0226, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12840960919857025, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0225, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.09958021342754364, + "learning_rate": 2.549819836273345e-05, + "loss": 0.02, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.10737278312444687, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0226, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.11473958194255829, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0219, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.12188851088285446, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0216, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.1229034960269928, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0227, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.1316317766904831, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1230575442314148, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0218, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13473588228225708, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0221, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1293632537126541, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0223, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.11663107573986053, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0198, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.147003173828125, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0216, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.10070289671421051, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0204, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12855689227581024, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0222, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.10310211777687073, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0227, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.25780248641967773, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0204, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.11621664464473724, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0196, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.15240156650543213, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0232, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.1359490305185318, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0191, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.13072653114795685, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.0242, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.1296282559633255, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0214, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.12183726578950882, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0215, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.10602327436208725, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0222, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.12424091249704361, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0215, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.14287705719470978, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0217, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.10017845034599304, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0214, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.1265304833650589, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0214, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.15279516577720642, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0194, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.11116814613342285, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0197, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.14380785822868347, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0187, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.12522602081298828, + "learning_rate": 2.541224092500406e-05, + "loss": 0.023, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.12386767566204071, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.0197, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11278095096349716, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0205, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12119927257299423, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0206, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.14128951728343964, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0197, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.12576138973236084, + "learning_rate": 2.539735331078111e-05, + "loss": 0.024, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.13276934623718262, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0217, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.15223100781440735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0222, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.13918957114219666, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0211, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.10129881650209427, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0192, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.15048135817050934, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0191, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.11579657346010208, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.019, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.1882474422454834, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0241, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.12608706951141357, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.021, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.14131402969360352, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0234, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.11546603590250015, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0208, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.0989922359585762, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0219, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.13154740631580353, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0234, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.12450987845659256, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0219, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.15696841478347778, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0228, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12059786915779114, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0236, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.1058221235871315, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0234, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.11429761350154877, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0232, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.11841994524002075, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0193, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.12387857586145401, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0239, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.11453676968812943, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0224, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.15571972727775574, + "learning_rate": 2.533461009735874e-05, + "loss": 0.022, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.10241609811782837, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.022, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.11682361364364624, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0211, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.13623486459255219, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0215, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.12484022229909897, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0212, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.09105533361434937, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0223, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.1024160087108612, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0174, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.14108790457248688, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0215, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.1268249899148941, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0211, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12255556881427765, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0192, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.15904900431632996, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0259, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.1356613039970398, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.023, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.12410393357276917, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0223, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1291753053665161, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0221, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.109084852039814, + "learning_rate": 2.529258887162501e-05, + "loss": 0.0243, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.16595998406410217, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0201, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.14062269032001495, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0232, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.13198593258857727, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0214, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.10668856650590897, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0229, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.11266572028398514, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0233, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.12742944061756134, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0221, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.12420892715454102, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0215, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.1265600174665451, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0215, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.11914686858654022, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0222, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.12876294553279877, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0233, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.09800750762224197, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0219, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1273864060640335, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0238, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.13299106061458588, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0226, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.09396592527627945, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.12955208122730255, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0204, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.16461864113807678, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0211, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.11535970121622086, + "learning_rate": 2.524135719213194e-05, + "loss": 0.02, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.11860322207212448, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0212, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10968036204576492, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0207, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.12494095414876938, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.02, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.10398012399673462, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0205, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.12939494848251343, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0216, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.10996326059103012, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0199, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.13704033195972443, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0208, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.11454308778047562, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.0212, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.14186057448387146, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0226, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.12624409794807434, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0199, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.14392296969890594, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.15774424374103546, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0213, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.1352480947971344, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0192, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.10594955086708069, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0231, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.11280091851949692, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0202, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11683343350887299, + "learning_rate": 2.519293379145164e-05, + "loss": 0.022, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.11721377819776535, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0223, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.18298691511154175, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.0225, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12927964329719543, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0202, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.1413726508617401, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.021, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.1330733746290207, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0228, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12341104447841644, + "learning_rate": 2.517472387591168e-05, + "loss": 0.021, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.12199880182743073, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.0202, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.12259527295827866, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.0227, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.11682213097810745, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0246, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.12783272564411163, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0204, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.09663784503936768, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0228, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.13686567544937134, + "learning_rate": 2.515648614306516e-05, + "loss": 0.02, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.12164216488599777, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0234, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.1371183544397354, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0217, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.12031334638595581, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0203, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.1183433011174202, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.1265946477651596, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0224, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.12686678767204285, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0202, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.13161398470401764, + "learning_rate": 2.513517369661726e-05, + "loss": 0.0216, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.13780610263347626, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0205, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.14213374257087708, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0237, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.12532930076122284, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0215, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1276065707206726, + "learning_rate": 2.512297821731217e-05, + "loss": 0.02, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.08893107622861862, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12819649279117584, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0202, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.09707193821668625, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0202, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.511077043763402e-05, + "loss": 0.0208, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.10047636926174164, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0219, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.15637417137622833, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0218, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.15387912094593048, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0213, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.14876650273799896, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0222, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.13631325960159302, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0214, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.12577500939369202, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0232, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15896311402320862, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0237, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.16459719836711884, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0223, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11081109195947647, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0223, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.12327340245246887, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0219, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14812195301055908, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0228, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.11862681806087494, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0252, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.10684223473072052, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0227, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.1279568374156952, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0216, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.13341103494167328, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0213, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.11913862824440002, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0214, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.13493330776691437, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.0214, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.1272144317626953, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0209, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.11348695307970047, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0217, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.16526950895786285, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0198, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.14953579008579254, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0231, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.12980960309505463, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0226, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.16652929782867432, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0244, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1722235530614853, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.026, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.23629626631736755, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0221, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.12417787313461304, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.023, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.11768155544996262, + "learning_rate": 2.502804732645032e-05, + "loss": 0.0186, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.13731959462165833, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.023, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.18034201860427856, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.0201, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11240512877702713, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0217, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.10463474690914154, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0201, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.1273874044418335, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0194, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.12314468622207642, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0237, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.13170552253723145, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.021, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.12725704908370972, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0198, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.11761748790740967, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0198, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13019445538520813, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0237, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.09089519083499908, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.0202, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.13669925928115845, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0236, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.0891583263874054, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0189, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.12015428394079208, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.0214, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.12889859080314636, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0214, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.1092396154999733, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.022, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.1011805608868599, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0201, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.11652060598134995, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.019, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.12280789762735367, + "learning_rate": 2.496950169240971e-05, + "loss": 0.0229, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.1253744661808014, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0245, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.15854160487651825, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0257, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.11671199649572372, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0204, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.15195102989673615, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.11878954619169235, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0207, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.1035882979631424, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0195, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12105780094861984, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0191, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.13278989493846893, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0207, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.16576215624809265, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.0228, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.13104894757270813, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0197, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.14476071298122406, + "learning_rate": 2.49354818242132e-05, + "loss": 0.023, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.14116041362285614, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0204, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.133507639169693, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0213, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11147522926330566, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0192, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.1317354440689087, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0229, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.15885432064533234, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0214, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.0996033102273941, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0213, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.11059842258691788, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0181, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.1103382259607315, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0252, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.12147006392478943, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.1167205348610878, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0217, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.11838674545288086, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0223, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.1745499074459076, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0213, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.09892687201499939, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0213, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.17135944962501526, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0228, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.13930797576904297, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0211, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.10764696449041367, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0204, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11211943626403809, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0207, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1359340399503708, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0215, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.11819283664226532, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0213, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.1166793555021286, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0216, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.14411170780658722, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0206, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.1380920261144638, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0211, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.12901239097118378, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.0224, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.1171313151717186, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.0207, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.15169396996498108, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0225, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.2066943198442459, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0182, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1179543286561966, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0226, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.12273375689983368, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.0224, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.09459412097930908, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0177, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.10208278894424438, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0207, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.11947600543498993, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.0246, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10635806620121002, + "learning_rate": 2.483599695058918e-05, + "loss": 0.021, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.15165211260318756, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0242, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.10573819279670715, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0221, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.10153646767139435, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0212, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.12596352398395538, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.021, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.11259406059980392, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0214, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.09709548950195312, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0186, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1462256908416748, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0228, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.15480898320674896, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0245, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.10707002878189087, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0197, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.10963082313537598, + "learning_rate": 2.480475069778109e-05, + "loss": 0.019, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.13645771145820618, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0229, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.08208916336297989, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0194, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.11223483830690384, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0212, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.12233787775039673, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0228, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.12175288051366806, + "learning_rate": 2.478909963381241e-05, + "loss": 0.0204, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.14267505705356598, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0203, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.12649142742156982, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0217, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.11254285275936127, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0214, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.1454164832830429, + "learning_rate": 2.477656539969673e-05, + "loss": 0.0221, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.17177407443523407, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0212, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.1089746281504631, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0222, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.1525813490152359, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0236, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.09651731699705124, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0214, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.12710145115852356, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0233, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.15999987721443176, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0215, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1600969433784485, + "learning_rate": 2.475460191424065e-05, + "loss": 0.0201, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.11045356094837189, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0208, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.14192573726177216, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.0216, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.11450321227312088, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0202, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.13454747200012207, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.021, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.11518846452236176, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0194, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.10673099011182785, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0214, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.12131883949041367, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0203, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.18098120391368866, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0182, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.09981405735015869, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0214, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.09194957464933395, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0209, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.11133445054292679, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0227, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.14570456743240356, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0181, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.14878204464912415, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0213, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.1475936472415924, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0229, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.14091694355010986, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0229, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.10411489754915237, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.021, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.10015816241502762, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0211, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09339930862188339, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0206, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.13016946613788605, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0223, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.10682770609855652, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0258, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.13658446073532104, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0232, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.12742926180362701, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.13928014039993286, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0208, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.10824346542358398, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0199, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.11346063017845154, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.0225, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12237603217363358, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0226, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.12201637774705887, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0209, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.09991955012083054, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0239, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.15387621521949768, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.025, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.1588744819164276, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0192, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.1347019523382187, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0196, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.12746664881706238, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0212, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11522521078586578, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0197, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.09697534143924713, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0219, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12183485180139542, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0195, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.13119301199913025, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0195, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.10748504102230072, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0192, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.1617487072944641, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0224, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.11518653482198715, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0207, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.17646238207817078, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0233, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.13729365170001984, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.09777814894914627, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0207, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.117672860622406, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0254, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.09014946967363358, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0214, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.1273731142282486, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0223, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10932891070842743, + "learning_rate": 2.460937119575347e-05, + "loss": 0.0224, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.1672651767730713, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0205, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.0930962786078453, + "learning_rate": 2.46030216185615e-05, + "loss": 0.02, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.09708375483751297, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0192, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.13240690529346466, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0214, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.10310725122690201, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0204, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.13352994620800018, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0206, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.15277482569217682, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0222, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.14223086833953857, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0229, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12845414876937866, + "learning_rate": 2.458077513854331e-05, + "loss": 0.0221, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13195961713790894, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0232, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.1461530476808548, + "learning_rate": 2.457441244940383e-05, + "loss": 0.022, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13729079067707062, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0196, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.15615087747573853, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0213, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.12907880544662476, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0198, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.11320316791534424, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0225, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.14982816576957703, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0225, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.12824060022830963, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0196, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.11940523982048035, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0204, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946357995271683, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.0218, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.13296495378017426, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0229, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.1383914053440094, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0207, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.11321834474802017, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0202, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.1276783049106598, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0193, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1406320035457611, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0191, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.12060298770666122, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0195, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.08627335727214813, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0191, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1429424285888672, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0197, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.15133187174797058, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0212, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.11892743408679962, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0231, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.11871744692325592, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0211, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.105750173330307, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.0234, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10969540476799011, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.022, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.09607745707035065, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0173, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.11796857416629791, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0193, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.114372618496418, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.02, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.1164821982383728, + "learning_rate": 2.449463406055895e-05, + "loss": 0.023, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.13561253249645233, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.018, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.14956215023994446, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0245, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.12785622477531433, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0222, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.16085495054721832, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.021, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10792174935340881, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0209, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.10508404672145844, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0217, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.11471311002969742, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0202, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.14265185594558716, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0224, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.13569262623786926, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0196, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15635624527931213, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0205, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12042741477489471, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0179, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.09488346427679062, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0198, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.20503900945186615, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0199, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.13667535781860352, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13254515826702118, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0204, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.13879507780075073, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0233, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.11448678374290466, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0237, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1255042403936386, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0204, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11896912008523941, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0207, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.12624827027320862, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0194, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.09312308579683304, + "learning_rate": 2.442727178914008e-05, + "loss": 0.022, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12632645666599274, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0189, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.11468527466058731, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0182, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.09245705604553223, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0188, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.12113099545240402, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0191, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.13049040734767914, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0221, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.16979938745498657, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0211, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.11175577342510223, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0172, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.1191241443157196, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0206, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.16317729651927948, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.0213, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.10087717324495316, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0194, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.12195733189582825, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0253, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11880146712064743, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0222, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.14213789999485016, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0198, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.15624478459358215, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.0205, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09695806354284286, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0219, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11260177940130234, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0206, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.15394112467765808, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0219, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.11663617193698883, + "learning_rate": 2.436928131964383e-05, + "loss": 0.0214, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.12596042454242706, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0225, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.10310690104961395, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0203, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.12332364916801453, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0203, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.16257967054843903, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0234, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.13102053105831146, + "learning_rate": 2.435313188161029e-05, + "loss": 0.022, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.14736537635326385, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0199, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.14611469209194183, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0238, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.14594200253486633, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0212, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.16921328008174896, + "learning_rate": 2.434019954415699e-05, + "loss": 0.0236, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.1318638175725937, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0201, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.10869060456752777, + "learning_rate": 2.433372911848899e-05, + "loss": 0.019, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.1446760594844818, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.021, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.1298162341117859, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10307328402996063, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0197, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.11932921409606934, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0209, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.10986137390136719, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0227, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.12099824845790863, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0182, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.13432320952415466, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.0213, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.14018790423870087, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0199, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.13015322387218475, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0191, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.10281360149383545, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0215, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.09500012546777725, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0216, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.15131708979606628, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0225, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.08750247210264206, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0221, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.10143480449914932, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0214, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.12935201823711395, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0202, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.10703578591346741, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0207, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08649012446403503, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0184, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.11186458170413971, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.021, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.11181459575891495, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0211, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.11756999045610428, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0216, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.12061341106891632, + "learning_rate": 2.426561884159603e-05, + "loss": 0.019, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.13861820101737976, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0214, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.11545911431312561, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0195, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.14892302453517914, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0226, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.10354511439800262, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0197, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.15290679037570953, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0208, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.12185964733362198, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0205, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.10486926138401031, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0206, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.10959216207265854, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0198, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.11638635396957397, + "learning_rate": 2.423633359536448e-05, + "loss": 0.0227, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.13661672174930573, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0206, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.13071168959140778, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0189, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.16377446055412292, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0203, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.22221329808235168, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0185, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.11251460015773773, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0191, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.10873523354530334, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0228, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.13705629110336304, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0203, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.13980528712272644, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.10355913639068604, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0199, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.1282302737236023, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.0215, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.1221374049782753, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0192, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.18542620539665222, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0199, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1250954121351242, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0189, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12665164470672607, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0217, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.09464617818593979, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.0197, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.09657798707485199, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0194, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.11533467471599579, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0189, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.13647402822971344, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0194, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.13276426494121552, + "learning_rate": 2.41743228236291e-05, + "loss": 0.022, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.16899874806404114, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0209, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.14313359558582306, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0211, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.14507170021533966, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0218, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.13025349378585815, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0209, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.1666184365749359, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0217, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14891882240772247, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0205, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11392978578805923, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0223, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.12964752316474915, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.019, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11125730723142624, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0211, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.08378217369318008, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0203, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.14530996978282928, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0206, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.10187243670225143, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0188, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.09652446955442429, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0192, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.10552513599395752, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0212, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11044801771640778, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0215, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.12248426675796509, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0185, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.11164992302656174, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0198, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.11919678002595901, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0187, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1675145924091339, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0262, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.15451745688915253, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0226, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.0997307077050209, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0196, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.1272345781326294, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0214, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.1286081224679947, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0214, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12226564437150955, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0211, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.11795010417699814, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0216, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.12963736057281494, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0215, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.10415369272232056, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0215, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.11642655730247498, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0194, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.19006073474884033, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0216, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.16677096486091614, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.0235, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.15928123891353607, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0195, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.10270075500011444, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0197, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.12253808975219727, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0216, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14145877957344055, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0211, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1387409120798111, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0185, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.11433693021535873, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0204, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.14247572422027588, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0187, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.12203793972730637, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.018, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.15705470740795135, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0212, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.10648002475500107, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0182, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.11573450267314911, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0211, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.11960671842098236, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0245, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.14416365325450897, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0178, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.11234421283006668, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0205, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.1375260204076767, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.025, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.08686859905719757, + "learning_rate": 2.402315322351119e-05, + "loss": 0.018, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11352542787790298, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0198, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.13374295830726624, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0216, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.09358088672161102, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0202, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.13767337799072266, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0229, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.1405765861272812, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.02, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11317892372608185, + "learning_rate": 2.400332803040664e-05, + "loss": 0.02, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.12356667965650558, + "learning_rate": 2.400002143698994e-05, + "loss": 0.0225, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.12570610642433167, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0232, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.1461946964263916, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0201, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.10893318802118301, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0211, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.10969080775976181, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0235, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.17320629954338074, + "learning_rate": 2.398347822254704e-05, + "loss": 0.0221, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14296430349349976, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0206, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12349261343479156, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0182, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.11617975682020187, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0209, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.1574973613023758, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0222, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.157145157456398, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.0199, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.14698748290538788, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0203, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.12345635890960693, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0212, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.1370248794555664, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0194, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.1332298070192337, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0219, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.12525111436843872, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.0219, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.14905311167240143, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0211, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12063094973564148, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.0262, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.10361988842487335, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0195, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.09919667989015579, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0201, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.11141914129257202, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.0223, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.17785966396331787, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0201, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.126350998878479, + "learning_rate": 2.392710390356112e-05, + "loss": 0.0201, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.12107345461845398, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0213, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.11466097086668015, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0202, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.15397872030735016, + "learning_rate": 2.391713512040237e-05, + "loss": 0.022, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.14412778615951538, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0223, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.10495682060718536, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0194, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.13021183013916016, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0199, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.13993357121944427, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0182, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.128571018576622, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.0181, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.11090676486492157, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0204, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.10988310724496841, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0203, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.11705116927623749, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0214, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.1277252584695816, + "learning_rate": 2.388719222953459e-05, + "loss": 0.0225, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1561414748430252, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0221, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.15278629958629608, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.02, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14865094423294067, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0213, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.1127776950597763, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0211, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.15085943043231964, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0212, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14555132389068604, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0216, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.14708220958709717, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0209, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.1093590185046196, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0199, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.1303861141204834, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0242, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.11462346464395523, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0197, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.12198365479707718, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0206, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.11193065345287323, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0201, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.13254034519195557, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0181, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.13355755805969238, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0204, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12551282346248627, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0212, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1441735178232193, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0217, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.11511780321598053, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0218, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.10427968204021454, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0222, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.11364301294088364, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0166, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.1401003897190094, + "learning_rate": 2.38204570134882e-05, + "loss": 0.0215, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.11645112931728363, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0198, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1340353935956955, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0212, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.0982050895690918, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0208, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.16713382303714752, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0185, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.13803045451641083, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.021, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.15552546083927155, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0233, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.10836438834667206, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0211, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.1831241250038147, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0198, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.12369874119758606, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0193, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.10911241918802261, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0221, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.11441849172115326, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0236, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.11105310916900635, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0206, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.13617953658103943, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0189, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.11214766651391983, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0192, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.12536488473415375, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0206, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.10356712341308594, + "learning_rate": 2.376687582517292e-05, + "loss": 0.019, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.13332195580005646, + "learning_rate": 2.376352133090187e-05, + "loss": 0.0216, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.13996818661689758, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0216, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.14305290579795837, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0204, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.10494769364595413, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0202, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.1461218297481537, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.0197, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.12469783425331116, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0207, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.13164709508419037, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.0207, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.1553804576396942, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0228, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1445176750421524, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0217, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12673451006412506, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0164, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.17502479255199432, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0208, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.11328406631946564, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0196, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12234057486057281, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0193, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.1333971917629242, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0214, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.11227905005216599, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0183, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.1287100911140442, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0198, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.10215240716934204, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0194, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.09786848723888397, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0202, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.11103200167417526, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0219, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.1491294801235199, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0202, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.1364760547876358, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.0198, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.11583423614501953, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0206, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.13896186649799347, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0183, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.0970211774110794, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.019, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.10326618701219559, + "learning_rate": 2.368281439559657e-05, + "loss": 0.022, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.11581890285015106, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0234, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10186298936605453, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0205, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11655393242835999, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0193, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.11990638077259064, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0217, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.10573728382587433, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0177, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.09257248789072037, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.0216, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.11310117691755295, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0188, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.12632012367248535, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0217, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1535339206457138, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0209, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.13133156299591064, + "learning_rate": 2.364907420469738e-05, + "loss": 0.0207, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11225332319736481, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0217, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.10744500905275345, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0212, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.10509403049945831, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0208, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.16215650737285614, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0199, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1208849623799324, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0196, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.14882729947566986, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0229, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.18775896728038788, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0206, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.14199234545230865, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0201, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.1510499119758606, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.021, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.16389605402946472, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0197, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.13032132387161255, + "learning_rate": 2.361188413960337e-05, + "loss": 0.0211, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.13895465433597565, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0222, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.10914628207683563, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0206, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.12215261161327362, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0186, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12954379618167877, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.0193, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.12779726088047028, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.02, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.11825352907180786, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0214, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11001764982938766, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0211, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.10464788973331451, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.02, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.0892137810587883, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0198, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.09811265021562576, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0191, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.14199955761432648, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0196, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.13846102356910706, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0216, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.12102426588535309, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0214, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.12360251694917679, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0199, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.12587665021419525, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.15562282502651215, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.0214, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.11366713047027588, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0192, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.15177540481090546, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0204, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.08908293396234512, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0186, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.1349143236875534, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0197, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.09388674795627594, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0205, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.11973906308412552, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0198, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.1290522664785385, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.02, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.12087979167699814, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0207, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.10237318277359009, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0189, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.10426577180624008, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0204, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13028934597969055, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0197, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.10696206241846085, + "learning_rate": 2.351686222335574e-05, + "loss": 0.0206, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.11024264991283417, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0189, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.11294405907392502, + "learning_rate": 2.35100554725415e-05, + "loss": 0.0228, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.1188240721821785, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0204, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.17485342919826508, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0202, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12582673132419586, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0209, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.18180538713932037, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0202, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.1459868997335434, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0189, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.09818759560585022, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0195, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.1473611295223236, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0253, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.09137327969074249, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0209, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.14026956260204315, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0202, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.1511334776878357, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0182, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.15016451478004456, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0204, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.14407901465892792, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.0205, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.1332983374595642, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0192, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.14312349259853363, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.0195, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.11917808651924133, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0194, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.17896629869937897, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0203, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.12084568291902542, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0199, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.14310620725154877, + "learning_rate": 2.344867863274048e-05, + "loss": 0.0201, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1168353408575058, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0225, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.14718633890151978, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0204, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.14102597534656525, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0197, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.09543738514184952, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0218, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1074792668223381, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0198, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.11555791646242142, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0196, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.12645067274570465, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0191, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.11634501069784164, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.02, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.09201087802648544, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0204, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.1155405044555664, + "learning_rate": 2.341449046471072e-05, + "loss": 0.0215, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.09871169179677963, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0216, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.11332481354475021, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0179, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.12474322319030762, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.0177, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.11339092999696732, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0221, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.1200975552201271, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.0209, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.14456698298454285, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0211, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.1371302306652069, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0165, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.11727641522884369, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0232, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.11336934566497803, + "learning_rate": 2.338366646946928e-05, + "loss": 0.0202, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.10325363278388977, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0209, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.11268003284931183, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0187, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10260385274887085, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.0195, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.10920321941375732, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0225, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.09120934456586838, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0194, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11567433178424835, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0197, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.10144805908203125, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0206, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1246950775384903, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0215, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.100107841193676, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0197, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.15822021663188934, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0202, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13518433272838593, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.02, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12268074601888657, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0196, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.12333078682422638, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0184, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.13588106632232666, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0218, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.12120439857244492, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0197, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.16428445279598236, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0198, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.11615289747714996, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.0182, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.1432264745235443, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0207, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.11717039346694946, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0198, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.10115417838096619, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0193, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.11894490569829941, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0198, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.11940047889947891, + "learning_rate": 2.330810220141532e-05, + "loss": 0.021, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.10992413759231567, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0195, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.10544339567422867, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0182, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.11725642532110214, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.0198, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.09130900353193283, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0209, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.09033828228712082, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0177, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.11721840500831604, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0207, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.12251118570566177, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0215, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.1496249884366989, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0219, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.14808881282806396, + "learning_rate": 2.327710137160376e-05, + "loss": 0.0212, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.145681232213974, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.019, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.12777844071388245, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0173, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.1110587939620018, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0198, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.11348012834787369, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0207, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.11685867607593536, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.02, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.17935989797115326, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0211, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.12398742884397507, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0191, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.10836582630872726, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0196, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.12038073688745499, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0207, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.12211567908525467, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0206, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.1702142357826233, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0227, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12037540227174759, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0191, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.15964926779270172, + "learning_rate": 2.323223255665661e-05, + "loss": 0.018, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.12493535876274109, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0212, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.1359032839536667, + "learning_rate": 2.32253202698687e-05, + "loss": 0.0219, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.18753379583358765, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0228, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.1336824893951416, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0201, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.09901099652051926, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0221, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.12925978004932404, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0206, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.1278320550918579, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.0178, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.12977559864521027, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0176, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10098735243082047, + "learning_rate": 2.320110760314603e-05, + "loss": 0.023, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.12594632804393768, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.0209, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.1140967383980751, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0181, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.1287144273519516, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0195, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.12329071760177612, + "learning_rate": 2.318725808577064e-05, + "loss": 0.0198, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.1339070051908493, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10581240057945251, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0265, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.13716311752796173, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0213, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.14610108733177185, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0195, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.128855898976326, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0185, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.0950121283531189, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0223, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.14444753527641296, + "learning_rate": 2.316299750241968e-05, + "loss": 0.017, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.12585613131523132, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0177, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.1290416419506073, + "learning_rate": 2.315606032389525e-05, + "loss": 0.02, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.10740494728088379, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0191, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11578185111284256, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0211, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.11840540170669556, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0209, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.09531640261411667, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0202, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.10436122864484787, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0188, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.14102807641029358, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0205, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14070506393909454, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0186, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.09883972257375717, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0196, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11524099856615067, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0187, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.11240902543067932, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0176, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.08062270283699036, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.019, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.13965697586536407, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0202, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.11376950889825821, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0218, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.13540907204151154, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0188, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.10888075083494186, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0216, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.10280206799507141, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0178, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.11512117087841034, + "learning_rate": 2.309699448641602e-05, + "loss": 0.02, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.09106973558664322, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0182, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10531709343194962, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0204, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.159554585814476, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0173, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.12161346524953842, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0196, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.08224594593048096, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0175, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.12767191231250763, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0234, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.12864384055137634, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0224, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.10232329368591309, + "learning_rate": 2.3069137227667e-05, + "loss": 0.0208, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.09609836339950562, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0183, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.1301981806755066, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0202, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.1258714497089386, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0217, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.12543867528438568, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13326771557331085, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.022, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.11770334839820862, + "learning_rate": 2.304821853869955e-05, + "loss": 0.019, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.11729256063699722, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0187, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.11887779086828232, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0224, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.10921293497085571, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0192, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.1346455216407776, + "learning_rate": 2.303426051834282e-05, + "loss": 0.022, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.1177918091416359, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0209, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.12912803888320923, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0191, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.12571996450424194, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0211, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.13959148526191711, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0198, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.1404334306716919, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0211, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.1111592948436737, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0213, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.10322082787752151, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0203, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.11235317587852478, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0187, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.13599546253681183, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0205, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11067865788936615, + "learning_rate": 2.299932279415038e-05, + "loss": 0.02, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.14130094647407532, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0204, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.14370687305927277, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0199, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.13825304806232452, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0197, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.11885961890220642, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.02, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.12376029789447784, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0199, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.10103051364421844, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0181, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.10931621491909027, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0198, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.11835317313671112, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0202, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.08036374300718307, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0189, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.08135218918323517, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0196, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.12611854076385498, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0211, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.09217566251754761, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0181, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.10258878767490387, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0194, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.1165921688079834, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0196, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.09290144592523575, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.0205, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.11575523763895035, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0204, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13866478204727173, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0178, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.13036414980888367, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0191, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11062757670879364, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0193, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.09973529726266861, + "learning_rate": 2.292926536293724e-05, + "loss": 0.02, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.13707594573497772, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0186, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11635670065879822, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0185, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.13926011323928833, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0192, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.13966591656208038, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0194, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.09878037869930267, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0191, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.11935862898826599, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0193, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.15291833877563477, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0189, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.12976837158203125, + "learning_rate": 2.290117482755701e-05, + "loss": 0.017, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.13367557525634766, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0178, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12121888250112534, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0191, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.13135644793510437, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0226, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.1142699271440506, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.017, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.09670589119195938, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0199, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.12751951813697815, + "learning_rate": 2.288008171679078e-05, + "loss": 0.018, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.14463315904140472, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0184, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.11006469279527664, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0198, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.15361396968364716, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0186, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.20782725512981415, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0208, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.15324300527572632, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.018, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.17235179245471954, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0218, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.12993113696575165, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0198, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.1496446281671524, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0203, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.16271376609802246, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.0215, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11106697469949722, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0206, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.10508760809898376, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0218, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.11433881521224976, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0203, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.12877102196216583, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0203, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.13301214575767517, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0225, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.15351711213588715, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0213, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.1114102303981781, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0212, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.09978143125772476, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0198, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10524465143680573, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0183, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.09780002385377884, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0185, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.1361900269985199, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0218, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.1329333484172821, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0191, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.10501913726329803, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.019, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.1329827904701233, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0198, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.10849577933549881, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0213, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.14308017492294312, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.0199, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.12313494831323624, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0193, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.10307908058166504, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.0215, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10591195523738861, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0178, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.1411430686712265, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0237, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13116225600242615, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0203, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.12032114714384079, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0201, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.12519164383411407, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0177, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.10340812057256699, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0194, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.10876207798719406, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0207, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11245796829462051, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0207, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11444602906703949, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0171, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12232744693756104, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0196, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.10049490630626678, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0179, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.09973384439945221, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.022, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.11037951707839966, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0197, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.14473563432693481, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0207, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.1138916090130806, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0198, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.12471994757652283, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0182, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.12418543547391891, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0195, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.12200118601322174, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.0217, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.13389603793621063, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0194, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10176926106214523, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.0189, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.10195910185575485, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0202, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.13289214670658112, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.018, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.10307228565216064, + "learning_rate": 2.270347347405331e-05, + "loss": 0.0183, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.149918794631958, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0172, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.1182866096496582, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0173, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.1035924181342125, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0194, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.108797088265419, + "learning_rate": 2.268928118647676e-05, + "loss": 0.0174, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.11457139253616333, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0208, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.10256518423557281, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0188, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13053470849990845, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0209, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.11632310599088669, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0187, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.12769649922847748, + "learning_rate": 2.267152769015067e-05, + "loss": 0.0206, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.13292407989501953, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0191, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.09822595119476318, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0189, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.13729946315288544, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0229, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.15705017745494843, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0203, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.12355367839336395, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0201, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.1278308629989624, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0205, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.09874208271503448, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0218, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.1316995620727539, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0212, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.10361800342798233, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0209, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.09672107547521591, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0205, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.12609392404556274, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0205, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.1360992044210434, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0188, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.11627763509750366, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0206, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.11733578145503998, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.02, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.1705910563468933, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0174, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.11704754829406738, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0199, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.13247141242027283, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0224, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.10916013270616531, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0202, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.11092681437730789, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.0186, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.12811405956745148, + "learning_rate": 2.260036839187291e-05, + "loss": 0.02, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.11893392354249954, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0194, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.16553092002868652, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0189, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.12431131303310394, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.02, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10850611329078674, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0222, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.12647055089473724, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0193, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.10432784259319305, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0196, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.1159503236413002, + "learning_rate": 2.257540798155951e-05, + "loss": 0.0192, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.10592516511678696, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0219, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.1318269520998001, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0195, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.11932560801506042, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0194, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.16499531269073486, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.02, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.1334158480167389, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.0217, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.14781776070594788, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0206, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.14816460013389587, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0206, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.11319883167743683, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0196, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.12413972616195679, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0206, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.1277540624141693, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0211, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.12178509682416916, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0192, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.1210661232471466, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0232, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.11300696432590485, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0182, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.15876832604408264, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0194, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.1387646645307541, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0202, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.12502169609069824, + "learning_rate": 2.251824990920905e-05, + "loss": 0.02, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.12365051358938217, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0197, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.12610606849193573, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0194, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.13096576929092407, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0208, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.13002553582191467, + "learning_rate": 2.250393750922939e-05, + "loss": 0.02, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.10048994421958923, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0193, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.11463778465986252, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0179, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10637608915567398, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0182, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.11959703266620636, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0183, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.11852603405714035, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12075617164373398, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0186, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.09626025706529617, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0195, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12773670256137848, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.019, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13364142179489136, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.019, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.14562146365642548, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0207, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.11752697080373764, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.13970085978507996, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0219, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.12456551194190979, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0203, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.36701101064682007, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0191, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.17254851758480072, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0193, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.1397981345653534, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0205, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.1273665875196457, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0227, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.08652572333812714, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0194, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.118954136967659, + "learning_rate": 2.243582938893974e-05, + "loss": 0.019, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.12105695903301239, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0173, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.10059279948472977, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0189, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.1476186215877533, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0192, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.11198431998491287, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0212, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.14829197525978088, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0216, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.14479608833789825, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0225, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.16737256944179535, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0202, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.12101127207279205, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0198, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.09408389776945114, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.019, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.10649813711643219, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0191, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.1328127533197403, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.02, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1135721281170845, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0181, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.14464536309242249, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0167, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.15146523714065552, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0216, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11212945729494095, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0189, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.11722676455974579, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0185, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10031717270612717, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0181, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.14852827787399292, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.02, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12114712595939636, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0207, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.12325385212898254, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0185, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.12915199995040894, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0188, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.10423776507377625, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0195, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.13398106396198273, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0182, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.1426725685596466, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0214, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.12515439093112946, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0184, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.15888792276382446, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0183, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.11037078499794006, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0199, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.10706029832363129, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.0195, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.11116144061088562, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0202, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.12716875970363617, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0182, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.1446460634469986, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0172, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.09947994351387024, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0205, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.09251125156879425, + "learning_rate": 2.231705291760663e-05, + "loss": 0.018, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.08440345525741577, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0181, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12813445925712585, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0206, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.1017315611243248, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0192, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.13247326016426086, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0212, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.14845387637615204, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0194, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.13824601471424103, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0224, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.1151830404996872, + "learning_rate": 2.229177978101565e-05, + "loss": 0.0176, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.12420724332332611, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0179, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.13085328042507172, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0179, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.12436439841985703, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.11873050034046173, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0197, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.1338828206062317, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0212, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.1106492280960083, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0179, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.16542823612689972, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0202, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.08829429000616074, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0176, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.10792568325996399, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0186, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.1279839128255844, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0221, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.10742732882499695, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0217, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.10220900177955627, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0201, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.1660427749156952, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.019, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.13747647404670715, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0225, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.13351675868034363, + "learning_rate": 2.223753180973339e-05, + "loss": 0.021, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.14151513576507568, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.0195, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.12823058664798737, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0189, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.12779751420021057, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0177, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.2243196815252304, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0183, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.12365520745515823, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0188, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.10981941968202591, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0191, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.11569637805223465, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0195, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.15785852074623108, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0222, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11610189825296402, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0172, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.1124223843216896, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0184, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.14634257555007935, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.0196, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.13306891918182373, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0204, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.08198657631874084, + "learning_rate": 2.219041683306619e-05, + "loss": 0.0177, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.09529033303260803, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0166, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.11170824617147446, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0195, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13825464248657227, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0191, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.11621235311031342, + "learning_rate": 2.217590132288247e-05, + "loss": 0.0202, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11921709030866623, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.019, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12485305964946747, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0205, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.1244436725974083, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0204, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11240405589342117, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0197, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.1087997704744339, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0183, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.11403695493936539, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0172, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.12764202058315277, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0161, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.17418256402015686, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0171, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.1522650271654129, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0197, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.16773882508277893, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0214, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.14681261777877808, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0208, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.14842623472213745, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.021, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.10614728182554245, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0182, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.15053948760032654, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0209, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.1381484866142273, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.0176, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.13161014020442963, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0174, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.09695997089147568, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0202, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.08903522044420242, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0171, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11456362903118134, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0183, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.1260307878255844, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.02, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.11732659488916397, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0197, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.126072496175766, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0196, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.11586430668830872, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.02, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.10504680871963501, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0193, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.09652367234230042, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0186, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10370496660470963, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0176, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.1192321851849556, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0178, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.10527302324771881, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0182, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10162831097841263, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0191, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.16128556430339813, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0203, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.11897512525320053, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0204, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.14880484342575073, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0205, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.11369120329618454, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.0193, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.11618658900260925, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0176, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.10052158683538437, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.0207, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1014244481921196, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0205, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.15118759870529175, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0215, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.09245041012763977, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0176, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.13509149849414825, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0224, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.12747371196746826, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0185, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11445175856351852, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0208, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11949758231639862, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0193, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.12797918915748596, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0182, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0185, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.11956185847520828, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0189, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.11755053699016571, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0184, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.1273607611656189, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0205, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.11855774372816086, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0188, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.09031623601913452, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0183, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12208469212055206, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0199, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.1083436831831932, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0193, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12228160351514816, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0182, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.1007765457034111, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0195, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.11114434152841568, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0173, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.1096438392996788, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0199, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.124867744743824, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0219, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.10032925009727478, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.0194, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.13164062798023224, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0193, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.1376691609621048, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0187, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.11254525184631348, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0186, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.12835443019866943, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.15313562750816345, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0219, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.16989216208457947, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0172, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.11433102935552597, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.019, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09817333519458771, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0222, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.09500890970230103, + "learning_rate": 2.193514973491427e-05, + "loss": 0.0176, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.14314892888069153, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0199, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.11764178425073624, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0169, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.12384455651044846, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0193, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.11438048630952835, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0174, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.13921929895877838, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0182, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.16880372166633606, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0174, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.1554516702890396, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0194, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.15355393290519714, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0212, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.1179167702794075, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0195, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.13388752937316895, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0208, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.10997489839792252, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0181, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.11783649772405624, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0189, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13451549410820007, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0186, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.1339322179555893, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0179, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.09972941130399704, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0214, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.10101358592510223, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0176, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.09771978110074997, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0177, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.14234617352485657, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0184, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.10462982952594757, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0197, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.09772231429815292, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0226, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.12018870562314987, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0164, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.10980863869190216, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0199, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.11285397410392761, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0172, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.1067347601056099, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0183, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.1046559289097786, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.0198, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.11118492484092712, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0179, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.1001342311501503, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0166, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.1053142324090004, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0188, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.16253869235515594, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.018, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.10042452812194824, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0173, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.11815430223941803, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0181, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.07585619390010834, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0195, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.17831926047801971, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0183, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.09953836351633072, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0194, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.12129344046115875, + "learning_rate": 2.18065424687869e-05, + "loss": 0.0188, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.09982063621282578, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0196, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.0906965360045433, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0172, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.1256277859210968, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.019, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.10290617495775223, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0171, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.17007306218147278, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0178, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.0779242068529129, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0192, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.12276264280080795, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0233, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.13211409747600555, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0184, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11407056450843811, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0211, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.13331526517868042, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0193, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.1603303849697113, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0192, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.10634022206068039, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0174, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.1658177524805069, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0207, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.1432579606771469, + "learning_rate": 2.175492159090619e-05, + "loss": 0.02, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.11264102905988693, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0183, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.12988954782485962, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.0199, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.1687532663345337, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0175, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12261682003736496, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0148, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.13096439838409424, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0175, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.13871102035045624, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0198, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.12536168098449707, + "learning_rate": 2.172907339903944e-05, + "loss": 0.018, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.12913931906223297, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0194, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.1255621761083603, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0172, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.12288706004619598, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0206, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.113137386739254, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0201, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.11107201129198074, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.14609375596046448, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.02, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.15227735042572021, + "learning_rate": 2.170320016677318e-05, + "loss": 0.016, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.12586566805839539, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0194, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.13495711982250214, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0187, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.11436781287193298, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.018, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.12271122634410858, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0193, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.16221405565738678, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0195, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.08645932376384735, + "learning_rate": 2.168100325341058e-05, + "loss": 0.0157, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.11750698834657669, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0186, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.1191735714673996, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0178, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.10067173838615417, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0187, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.09761539846658707, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0169, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.13806267082691193, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0217, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.10618548095226288, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0165, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.15135695040225983, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0199, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10563432425260544, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0176, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.12757790088653564, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0183, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.11258719861507416, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.0162, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.15290242433547974, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0192, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.1350955218076706, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0188, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.09494426846504211, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0187, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12816165387630463, + "learning_rate": 2.162913952762737e-05, + "loss": 0.019, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.1237507089972496, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0187, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.12867869436740875, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0202, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11882596462965012, + "learning_rate": 2.161801300808089e-05, + "loss": 0.019, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.10291648656129837, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0187, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.09269602596759796, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.019, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.08865118771791458, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.0188, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10444393008947372, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0182, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.11461203545331955, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.0219, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.13728564977645874, + "learning_rate": 2.159574640654742e-05, + "loss": 0.0223, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.1394069343805313, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0172, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.15394151210784912, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0182, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.10422483086585999, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0175, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.10996434837579727, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0182, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.1026761382818222, + "learning_rate": 2.157717712784381e-05, + "loss": 0.018, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.1373908817768097, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0226, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.10155291855335236, + "learning_rate": 2.156974591793542e-05, + "loss": 0.0179, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.14904215931892395, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0192, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.1131916344165802, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0175, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.09735766798257828, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0182, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.13132493197917938, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0193, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.12934266030788422, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0184, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.1196565330028534, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0186, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.12375814467668533, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.0199, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.0892246812582016, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0178, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.0932748094201088, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0192, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11503782123327255, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0187, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.08250989019870758, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0183, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.101202592253685, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0183, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1226576566696167, + "learning_rate": 2.152139450890464e-05, + "loss": 0.022, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.1167614683508873, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0176, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.12396091967821121, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0186, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.13461633026599884, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.0169, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.10792742669582367, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0189, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11649876832962036, + "learning_rate": 2.150277549265773e-05, + "loss": 0.02, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.1314917951822281, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0157, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.14143578708171844, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0197, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10042720288038254, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0177, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.12019779533147812, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0164, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.14096254110336304, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0223, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.10627058893442154, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0171, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.11092417687177658, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.019, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.12211407721042633, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0199, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11551571637392044, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0197, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.10474767535924911, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.018, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.12364303320646286, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0178, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13243450224399567, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0197, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10594736039638519, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0179, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.11143538355827332, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0196, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.10518378764390945, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0197, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12268701940774918, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0176, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.09408178925514221, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0205, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.12529458105564117, + "learning_rate": 2.143564506261267e-05, + "loss": 0.0184, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.15819092094898224, + "learning_rate": 2.143191093594534e-05, + "loss": 0.021, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.15818199515342712, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0182, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.13374921679496765, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0177, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.10046645253896713, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0197, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.11422070860862732, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0185, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.1138019785284996, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0182, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.09617039561271667, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0178, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.1452113538980484, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0222, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.10409227013587952, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0196, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.11277221143245697, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0175, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.12325838208198547, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0201, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12721778452396393, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0174, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.14016005396842957, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.017, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.11149191111326218, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0168, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.13445520401000977, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0194, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10086208581924438, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0197, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.115168996155262, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0181, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.11483300477266312, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.0183, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.1196938082575798, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0172, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.12219196557998657, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0177, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.11318044364452362, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.0204, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.09219531714916229, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0181, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.09911682456731796, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0192, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.11822745203971863, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0192, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.11101230978965759, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0165, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.09978491812944412, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0203, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.09874484688043594, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0199, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.08682837337255478, + "learning_rate": 2.133090580749213e-05, + "loss": 0.0177, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.14352355897426605, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0189, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.11133426427841187, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0199, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.10667043924331665, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0183, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.14207594096660614, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0218, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.12165568023920059, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0167, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.09311263263225555, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0183, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10423581302165985, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0151, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.12640471756458282, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0214, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.11349574476480484, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0171, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.14359329640865326, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0182, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.10977838933467865, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0181, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.12443531304597855, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0168, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.12441008538007736, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0186, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.14103129506111145, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0197, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10460726171731949, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0178, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.12526848912239075, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0188, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.13363681733608246, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0183, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.13106584548950195, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0181, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.1353096216917038, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0202, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.11889169365167618, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0198, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.12628430128097534, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0178, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.13260240852832794, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0182, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12357809394598007, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0175, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.13913387060165405, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0189, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.12303008884191513, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0187, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.10231908410787582, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.017, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.20065239071846008, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0172, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.11311347782611847, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.019, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.131951242685318, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0182, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.15945149958133698, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0202, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09272884577512741, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0189, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.0937163308262825, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0208, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.10963460803031921, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0197, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.1618727743625641, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0195, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.0919048935174942, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0177, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.12060586363077164, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0209, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.13121892511844635, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0181, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.11795026063919067, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0186, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.13060589134693146, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0226, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.08142203092575073, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0181, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.12634120881557465, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0178, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.1448773890733719, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0179, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1230989471077919, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0188, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.11944684386253357, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0212, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.1445692479610443, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0184, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.0831015482544899, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0171, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.1456967294216156, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0189, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.11207464337348938, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0185, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.16137553751468658, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0182, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.09820980578660965, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0192, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.11823676526546478, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0183, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.12873874604701996, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0181, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.11243417114019394, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0188, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.11120739579200745, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0222, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.10704528540372849, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0178, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.12063440680503845, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0174, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.13275623321533203, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0171, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.10400331765413284, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0201, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.10485411435365677, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0187, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.12517869472503662, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0187, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.11703156679868698, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.0178, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.11982452869415283, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0186, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11858765780925751, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0182, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.10404367744922638, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0173, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.09697552025318146, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0188, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.12046534568071365, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0173, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.0896986722946167, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0188, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.10993069410324097, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0213, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.1200791671872139, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0179, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.14664512872695923, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.0202, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.09285914152860641, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0207, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.1080656424164772, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0188, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.140005961060524, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0188, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.11084222793579102, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.02, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.0913061872124672, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0194, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.09459970891475677, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0182, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.09709502756595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0195, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.11101861298084259, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0184, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.1152782216668129, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0178, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.11495838314294815, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0178, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.14250393211841583, + "learning_rate": 2.10258099183954e-05, + "loss": 0.019, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.09898682683706284, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0173, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.13904057443141937, + "learning_rate": 2.101823807728166e-05, + "loss": 0.018, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.12013959139585495, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0163, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.0925723984837532, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0187, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.1284961700439453, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0158, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.12408998608589172, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0206, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1095876470208168, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0172, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.10166801512241364, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0183, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13473083078861237, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0155, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.10492225736379623, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0188, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.10769215226173401, + "learning_rate": 2.098414220174082e-05, + "loss": 0.021, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.09331739693880081, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0176, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.09247364103794098, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0163, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.20951682329177856, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0203, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.16750548779964447, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0196, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.10933154821395874, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0169, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10767697542905807, + "learning_rate": 2.09613911545763e-05, + "loss": 0.0184, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.11565674096345901, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0208, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.11934554576873779, + "learning_rate": 2.095380384728948e-05, + "loss": 0.018, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.12786556780338287, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0169, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.11752485483884811, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0194, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.11586920917034149, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0179, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.08849998563528061, + "learning_rate": 2.0938623809196e-05, + "loss": 0.018, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.12535157799720764, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0176, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.10429015755653381, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0188, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.09504082053899765, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0186, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.12917572259902954, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0182, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.11084037274122238, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0163, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.12090355157852173, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0196, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.11049258708953857, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0194, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.0831870436668396, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0188, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11006250977516174, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.018, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.12107989192008972, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0178, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.10361241549253464, + "learning_rate": 2.089684155251222e-05, + "loss": 0.017, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.13428105413913727, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0198, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.13375632464885712, + "learning_rate": 2.088923894557553e-05, + "loss": 0.018, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.14561115205287933, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0181, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.09850531816482544, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0176, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.11646582931280136, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0162, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.14659325778484344, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0168, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.15971294045448303, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0198, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.11780509352684021, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0156, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11453750729560852, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0179, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.09094218909740448, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0172, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.13252301514148712, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0209, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.10068004578351974, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0159, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.10302501171827316, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0172, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.10893740504980087, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.019, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.12446500360965729, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0177, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1345142424106598, + "learning_rate": 2.083597073454921e-05, + "loss": 0.021, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.12400489300489426, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0194, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.15042100846767426, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0162, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.1559734046459198, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0189, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.10689649730920792, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0163, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.15904614329338074, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0199, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.11539188027381897, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0208, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.14550678431987762, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0191, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.10212639719247818, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0177, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.10318224877119064, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0176, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0196, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.10269734263420105, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0179, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.11070621758699417, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0196, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13299307227134705, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.02, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11430692672729492, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0188, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10718249529600143, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.0204, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.11855681240558624, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.019, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.10754020512104034, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0157, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11011523753404617, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0172, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.11931152641773224, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.0171, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.12425696104764938, + "learning_rate": 2.075972279799841e-05, + "loss": 0.0195, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.10292454063892365, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0177, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.11290059238672256, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0165, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.1453622728586197, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0207, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.12152253836393356, + "learning_rate": 2.07444521407078e-05, + "loss": 0.017, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.134286567568779, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0198, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.11180262267589569, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0196, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.12045837193727493, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.018, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.18509334325790405, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0209, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.17120368778705597, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0197, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.11996874213218689, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0204, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.12391912192106247, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0201, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.1248779147863388, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.016, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.12873846292495728, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0184, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.11072682589292526, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.0182, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.10367089509963989, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0179, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.12854290008544922, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0184, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.16348853707313538, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0169, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.1345866173505783, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0187, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.1139485239982605, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0178, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.1656051129102707, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0177, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.1133129820227623, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0154, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.0865294560790062, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0163, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.1368715763092041, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0218, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.11601361632347107, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0185, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.1401238590478897, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0177, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.10183753073215485, + "learning_rate": 2.066033925513678e-05, + "loss": 0.0192, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.10317160934209824, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0153, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.12317138910293579, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.021, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.10953490436077118, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0191, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.10530596971511841, + "learning_rate": 2.064502355956822e-05, + "loss": 0.0179, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.09054628014564514, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0189, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.12472135573625565, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0239, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.1551167368888855, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0189, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.08141466975212097, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0174, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.13025878369808197, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0185, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.13056428730487823, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0183, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.1187913790345192, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0182, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.11957527697086334, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0173, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.09339714050292969, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0182, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.1584015190601349, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.02, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.13754191994667053, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0193, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.09880727529525757, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0192, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.15097497403621674, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0172, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12028415501117706, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0182, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.10583297163248062, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0183, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.09343970566987991, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0187, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.1436399519443512, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0175, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.10790709406137466, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0184, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.14520230889320374, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0187, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.10826292634010315, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0188, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.1122521162033081, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0197, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.11445961147546768, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.015, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.1473265290260315, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0193, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.1572754830121994, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0186, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.1085842102766037, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0169, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.11982112377882004, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0163, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.12014015763998032, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.0217, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.11440391838550568, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0175, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.1276320368051529, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0203, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12434501200914383, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0186, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.11981593072414398, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0138, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.08389589190483093, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0162, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.11693917959928513, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.0199, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.1033058762550354, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0182, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.17878879606723785, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0181, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.12847506999969482, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0186, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.10514750331640244, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0166, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.1037413701415062, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0182, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.10402073711156845, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0165, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.10330315679311752, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0164, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.14082680642604828, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0205, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.1187078058719635, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0178, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.10590347647666931, + "learning_rate": 2.047995045386578e-05, + "loss": 0.0192, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.11178472638130188, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.019, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.11713241040706635, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0169, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.1520795077085495, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0173, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.11906655877828598, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0163, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.10854434967041016, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0181, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.16785599291324615, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.0212, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10305774211883545, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0179, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.09346644580364227, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0162, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11048682034015656, + "learning_rate": 2.044530211206255e-05, + "loss": 0.0158, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.14503270387649536, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0185, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.12631897628307343, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0182, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.14716260135173798, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0205, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.1364247053861618, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0187, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.09775770455598831, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0179, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.17999885976314545, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0194, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.1381615400314331, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0192, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.10749474912881851, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0189, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.15220865607261658, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.12083570659160614, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0175, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.11410468816757202, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0174, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.11285928636789322, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.0191, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.09808669239282608, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0173, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.07822877913713455, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0179, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.1032382920384407, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0177, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12453432381153107, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0183, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.1397278904914856, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0175, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.1701551228761673, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0176, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.13178907334804535, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0158, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.16403238475322723, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0182, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.13050618767738342, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.019, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09723208099603653, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.016, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.11937404423952103, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0194, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.10595990717411041, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0179, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.10598801076412201, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0172, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.12562280893325806, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0192, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.1062425896525383, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0167, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.19852802157402039, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0186, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10752324014902115, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0171, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.11313100904226303, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0191, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.09881386160850525, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0189, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.11926184594631195, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0176, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.11304130405187607, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0171, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.159687340259552, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0209, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.12324368953704834, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0184, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.10718217492103577, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0178, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12370229512453079, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0181, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.11843078583478928, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.017, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.11866020411252975, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0173, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.10999849438667297, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0217, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.10843008756637573, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.0163, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.13494807481765747, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0202, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.13634233176708221, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.0203, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.09910089522600174, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0154, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.1105995923280716, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0172, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.11397486180067062, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0171, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10141155868768692, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0187, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.09431210905313492, + "learning_rate": 2.025995044916193e-05, + "loss": 0.0163, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.09627058357000351, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0188, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.10498029738664627, + "learning_rate": 2.025220726516947e-05, + "loss": 0.0197, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.12692295014858246, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0163, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.11992908269166946, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0188, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.1448463350534439, + "learning_rate": 2.024058949838742e-05, + "loss": 0.0182, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.10764076560735703, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0191, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.1176103726029396, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0202, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.10393970459699631, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0174, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.12357908487319946, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0188, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11184806376695633, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0171, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09287009388208389, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0164, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.13145606219768524, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0175, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.108634814620018, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0192, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.1643940508365631, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.02, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.1642858386039734, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0195, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.11515326052904129, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0173, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.11438507586717606, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0187, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.1556805521249771, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0193, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.1081680878996849, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0176, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.11440131068229675, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0198, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.13953827321529388, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0163, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.12443852424621582, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0173, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.12884436547756195, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.0184, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.09170050173997879, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0158, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.08685769885778427, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0143, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.12880417704582214, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0172, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.09957600384950638, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0189, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.11359020322561264, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0191, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.10620398819446564, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0183, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.11106064170598984, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0175, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.14327868819236755, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0156, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.09743347764015198, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0165, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.11041118204593658, + "learning_rate": 2.013198534613597e-05, + "loss": 0.019, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.13757234811782837, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.017, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.10167638957500458, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0172, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.1386769413948059, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0157, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.13493874669075012, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0184, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.1238163560628891, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0178, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12875531613826752, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0176, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11199075728654861, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0179, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12402849644422531, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0197, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.12012362480163574, + "learning_rate": 2.009701166604228e-05, + "loss": 0.0203, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.1319619119167328, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0176, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.11875088512897491, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.017, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.14083963632583618, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0192, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1157609224319458, + "learning_rate": 2.008145773123833e-05, + "loss": 0.017, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1447380781173706, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0192, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.08668482303619385, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.0182, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.11417029052972794, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0162, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.12431302666664124, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0176, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.09466125816106796, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0174, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.13191643357276917, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0162, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.16682866215705872, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0167, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.09262735396623611, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0174, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10246162116527557, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0157, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.14533615112304688, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0193, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10985267907381058, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0185, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.1364467293024063, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0174, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.12792983651161194, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.019, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.13055621087551117, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0175, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.1272381991147995, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0191, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.1523309350013733, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.0192, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.12438428401947021, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0183, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.13417088985443115, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0192, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.10955093055963516, + "learning_rate": 2.000749254391884e-05, + "loss": 0.019, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.1140032708644867, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.0155, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.13144028186798096, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.0177, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.10048601031303406, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0169, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.1272922307252884, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0167, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.10301118344068527, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0166, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.10341407358646393, + "learning_rate": 1.998410649877165e-05, + "loss": 0.018, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.08479570597410202, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0193, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.10605745017528534, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0169, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.09806226938962936, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0175, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.12049946188926697, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0187, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.13633446395397186, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0167, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.12669390439987183, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.02, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.1395879089832306, + "learning_rate": 1.99568055617237e-05, + "loss": 0.02, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.1145915761590004, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.019, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12087064981460571, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0194, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.11252355575561523, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0156, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.10857690125703812, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0197, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.17054887115955353, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0172, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.12123709172010422, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0184, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.08856352418661118, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0178, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.09841718524694443, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0198, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.1236206591129303, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0171, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11047739535570145, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0198, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.11695606261491776, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0162, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.12013910710811615, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0171, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.08011548966169357, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0176, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.13469302654266357, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0177, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.12169095873832703, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.0192, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.1165102869272232, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0165, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.16436143219470978, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0164, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.11628631502389908, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0174, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.13279633224010468, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0164, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11530247330665588, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0199, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13026569783687592, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0193, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.12264525890350342, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0178, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.09320709109306335, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0157, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.11837475001811981, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.016, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.12293318659067154, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.0182, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.09163888543844223, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0197, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.1171371266245842, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0162, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.1096082329750061, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0198, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09364283084869385, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0164, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.08231621980667114, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0169, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.1273244321346283, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.0167, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.15531282126903534, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0194, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.13366153836250305, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0186, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.1072721853852272, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0173, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.11176268011331558, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0182, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.10074266046285629, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0167, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.11638782918453217, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0194, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.1109209880232811, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0197, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.1511884480714798, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0202, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.09859053790569305, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0185, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.12002220004796982, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0177, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.11930137127637863, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0159, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.10564088076353073, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0176, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.10428094118833542, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0197, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.1335987150669098, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0188, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.12865577638149261, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0179, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.14246998727321625, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0181, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.1250566691160202, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0188, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.09627831727266312, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0153, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.09008506685495377, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0185, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.12598916888237, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0169, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.11335116624832153, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0165, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.10588988661766052, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0171, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.10731689631938934, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.0174, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.11184538155794144, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0154, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.10870286822319031, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0163, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.11692065745592117, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0178, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.10036200284957886, + "learning_rate": 1.972989772246684e-05, + "loss": 0.016, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.12077154219150543, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0178, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.1272457093000412, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0188, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.08074570447206497, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0159, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.11381031572818756, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.0155, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.09836770594120026, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0161, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.09236133843660355, + "learning_rate": 1.970635454034925e-05, + "loss": 0.0183, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.1167517825961113, + "learning_rate": 1.970242942396647e-05, + "loss": 0.018, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.10953414440155029, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.015, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.12463351339101791, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0192, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.1285952925682068, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0172, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.10794857144355774, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0167, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.11815376579761505, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0176, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.10251157730817795, + "learning_rate": 1.967887123661664e-05, + "loss": 0.018, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.16941046714782715, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0182, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.14373992383480072, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.0177, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.11837214231491089, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0169, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.10599380731582642, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0167, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.11006144434213638, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0207, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.11699084937572479, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0164, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.10131341218948364, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0177, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.09975451976060867, + "learning_rate": 1.964744043284577e-05, + "loss": 0.017, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.08056963980197906, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0171, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.16518834233283997, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0186, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.10148414224386215, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0181, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.14867499470710754, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0183, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.16952896118164062, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0199, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13692231476306915, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0198, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.15133927762508392, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0184, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11434897035360336, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0189, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.0917418971657753, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0167, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12218283116817474, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0168, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.12193279713392258, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0173, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.09389539062976837, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0182, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.10455376654863358, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0169, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.11151516437530518, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0188, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.14512453973293304, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0194, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.10494859516620636, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0169, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.15839508175849915, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0177, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.12615326046943665, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0172, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.1135321632027626, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0166, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.12872013449668884, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0173, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.18429434299468994, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0165, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.13249747455120087, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0183, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.10603360831737518, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0145, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.10620120912790298, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0155, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.10986930131912231, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0163, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09298109263181686, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0179, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09945658594369888, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0178, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.08814677596092224, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0186, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.09042713046073914, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0173, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.11592321842908859, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0191, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.1026391014456749, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.018, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.09040834754705429, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0164, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.11391181498765945, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0177, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11113253980875015, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.0166, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.10178712755441666, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0167, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.13696520030498505, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0171, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.10290519148111343, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0164, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13472187519073486, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0173, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.08172356337308884, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0187, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.12297531962394714, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0166, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.13925355672836304, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0193, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.15394000709056854, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0169, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.09050899744033813, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.019, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.11854210495948792, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0181, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.10207716375589371, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0154, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.10006722062826157, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0161, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.10565274208784103, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0183, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.11656428128480911, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0166, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.12020296603441238, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0181, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.0964578241109848, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0174, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.0888240784406662, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0158, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.09271078556776047, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0187, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.10407985746860504, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0152, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1329568326473236, + "learning_rate": 1.94346993525268e-05, + "loss": 0.0166, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.13284213840961456, + "learning_rate": 1.943075029574398e-05, + "loss": 0.018, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.14158464968204498, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0194, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.11495806276798248, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0157, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.12886112928390503, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0174, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08954162895679474, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0158, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.09367296099662781, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0171, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.11188989877700806, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0162, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.10758768022060394, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0154, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.11783428490161896, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0182, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.18594273924827576, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0175, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.08010479807853699, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0167, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13949649035930634, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0175, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.10709669440984726, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0179, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.1675017923116684, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0176, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12764538824558258, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0165, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11073391884565353, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0169, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.10426445305347443, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0171, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.09000374376773834, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0161, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.10287991911172867, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0171, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.12322332710027695, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0178, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.11188322305679321, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0166, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.14056499302387238, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0196, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.08969348669052124, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0169, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.10100627690553665, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0159, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.1181580200791359, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0185, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.11281560361385345, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0171, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.11793895065784454, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0187, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.14534896612167358, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0177, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.10177207738161087, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0182, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.1335727721452713, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.017, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.14608687162399292, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0197, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.09763725847005844, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0184, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.09486788511276245, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0177, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.12235541641712189, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.018, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.13838450610637665, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0174, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.09837213158607483, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.0174, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.11713139712810516, + "learning_rate": 1.928836248385875e-05, + "loss": 0.016, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.13015268743038177, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0187, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.13498486578464508, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0181, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.12189928442239761, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0156, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.14399787783622742, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0185, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.10662715882062912, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0164, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.1145375519990921, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0185, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.11677944660186768, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0153, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.10242128372192383, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0168, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.1290024369955063, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0189, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.11820637434720993, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0167, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.12113048881292343, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0187, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.12122460454702377, + "learning_rate": 1.92408053936279e-05, + "loss": 0.0175, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.09413880109786987, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0149, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.11389796435832977, + "learning_rate": 1.923287468901755e-05, + "loss": 0.0185, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13923633098602295, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.0169, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.15947312116622925, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0188, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.14314432442188263, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0182, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.1426798403263092, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0179, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.12179753184318542, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.02, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11758018285036087, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0197, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11349253356456757, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0171, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.13724955916404724, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0181, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.1260363608598709, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0172, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.09541817009449005, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0181, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.0922083631157875, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0177, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.08051180839538574, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0167, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.11803245544433594, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0177, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10018756985664368, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.10512874275445938, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0164, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.17478078603744507, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0169, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.08864858746528625, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0165, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11313372850418091, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0172, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.09432164579629898, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0163, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.09435626119375229, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0188, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.10041314363479614, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0157, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.07982632517814636, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0176, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.09699293971061707, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0177, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.11059583723545074, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.017, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.11384644359350204, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0163, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.11473467200994492, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0165, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.11018316447734833, + "learning_rate": 1.912568530871305e-05, + "loss": 0.0179, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.10781396180391312, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0199, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.11702359467744827, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0165, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.12909658253192902, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0162, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.09839031100273132, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0152, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.1357898861169815, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0198, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.11221814155578613, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.0178, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.11431093513965607, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0176, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.11436092853546143, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0187, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.09838918596506119, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0148, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.11489785462617874, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0168, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.10876795649528503, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0169, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.1032189428806305, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0152, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.10013218969106674, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.11554815620183945, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0182, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.2529490292072296, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0198, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.09215852618217468, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0168, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.12875677645206451, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0185, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.1232401430606842, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0179, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.10334797203540802, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0171, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10524798929691315, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0162, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.11014945805072784, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0176, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.1074177473783493, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0162, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.14043818414211273, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0191, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.10011711716651917, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0178, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0172, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.12526027858257294, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0137, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.15774008631706238, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0153, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.1049765795469284, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0155, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09307108074426651, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0186, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1369069367647171, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0158, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.12320058792829514, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0181, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.10885081440210342, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0168, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.08989177644252777, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0153, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.09219451993703842, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0172, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.11661440879106522, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0192, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.0971342995762825, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0178, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.10453730076551437, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0165, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.17652514576911926, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0196, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10733854025602341, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0154, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.0989762544631958, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0181, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.12107723951339722, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.1385297179222107, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0191, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.12209209054708481, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0217, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.09332361072301865, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0158, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.12448447197675705, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0164, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.12061425298452377, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0169, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.12132656574249268, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0183, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.14007116854190826, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0167, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.11448334902524948, + "learning_rate": 1.893057716200171e-05, + "loss": 0.015, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.12702956795692444, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0152, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.09236849099397659, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0177, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.09818986803293228, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.0164, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12131274491548538, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0171, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10609138011932373, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.0147, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.1431477814912796, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0173, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.0844314768910408, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0162, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.09800038486719131, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0157, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.11529140919446945, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0172, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.086520716547966, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0154, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.12068255245685577, + "learning_rate": 1.888667760339568e-05, + "loss": 0.015, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.12278824299573898, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0174, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.10242115706205368, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0173, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.11696609109640121, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0187, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.11133866757154465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0151, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.1319858729839325, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0165, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.08355733752250671, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0166, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.10663259029388428, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0177, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.09787473827600479, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0154, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.12111647427082062, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0171, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.10036861151456833, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.15351398289203644, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.019, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.11561891436576843, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0174, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.09197179228067398, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0174, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.0961032509803772, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.019, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.09264974296092987, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0158, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.11056753247976303, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0159, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.10464371740818024, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0165, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.13888190686702728, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0185, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.09801805019378662, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0189, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.15648017823696136, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0167, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1129915863275528, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0166, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.10456246137619019, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0171, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.10873697698116302, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.0168, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.17470459640026093, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0169, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.15504027903079987, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.10821744054555893, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0152, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.14031334221363068, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0181, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.128667414188385, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0183, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.09095419198274612, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0151, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.09303424507379532, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0167, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.10706666111946106, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0174, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.13011431694030762, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0177, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.13821059465408325, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0154, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.10680826753377914, + "learning_rate": 1.875076403503638e-05, + "loss": 0.017, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.12847846746444702, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0146, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.1472046673297882, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0184, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10622651875019073, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0153, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.129075825214386, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0189, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.12048091739416122, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.017, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.10954522341489792, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0175, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.10811172425746918, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0173, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.13378627598285675, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0193, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.13185591995716095, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0195, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.1013568863272667, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0182, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.07558644562959671, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0173, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.10668148100376129, + "learning_rate": 1.870271530813327e-05, + "loss": 0.018, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.134077250957489, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0157, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.14325739443302155, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0221, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.1389995515346527, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0178, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.13043907284736633, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0169, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.12082460522651672, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0174, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11521385610103607, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0165, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.11550573259592056, + "learning_rate": 1.867466815081593e-05, + "loss": 0.017, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.11315014213323593, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0173, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.1067388728260994, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0161, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.0941673144698143, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0158, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.09685055166482925, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0161, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.11744765937328339, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0164, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.10552993416786194, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0154, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.12654545903205872, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0184, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.11333387345075607, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0186, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.09856864809989929, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.0187, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.0933985561132431, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0171, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.096091628074646, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0194, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.08463368564844131, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0165, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.10648614913225174, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.017, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.10691184550523758, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0165, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1005169153213501, + "learning_rate": 1.861452118736741e-05, + "loss": 0.0174, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.1037023589015007, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0172, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.1299487054347992, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.015, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14639562368392944, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0167, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.12270855903625488, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0173, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.09706515073776245, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0169, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.11888926476240158, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.0152, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.11289773136377335, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0175, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13422279059886932, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0168, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.10992032289505005, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0176, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.10719195753335953, + "learning_rate": 1.85743888359202e-05, + "loss": 0.017, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.10348568856716156, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0166, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.10436776280403137, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0172, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.09779159724712372, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.018, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.10799746960401535, + "learning_rate": 1.855832827763672e-05, + "loss": 0.016, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.11972185224294662, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0195, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.10282087326049805, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0169, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.11840159446001053, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0183, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.13861460983753204, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0205, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10774466395378113, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0163, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.14446309208869934, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0168, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.1313662976026535, + "learning_rate": 1.853021190846946e-05, + "loss": 0.0161, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.0857798308134079, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0159, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.11933837085962296, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0178, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11762946099042892, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0154, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.11128528416156769, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0198, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.07360108196735382, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0148, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.10680225491523743, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0174, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.11387799680233002, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0188, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.2774786651134491, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0165, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.12314122170209885, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.017, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.14403507113456726, + "learning_rate": 1.849002290456207e-05, + "loss": 0.0175, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.10811382532119751, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0173, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.1399172991514206, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0154, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.10577643662691116, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0177, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.10794928669929504, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0155, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.0844842791557312, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0149, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.09744124114513397, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0166, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.14749610424041748, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0179, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.09801949560642242, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0177, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.07795082777738571, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0174, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.11197780072689056, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.0172, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.11819472163915634, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.08461402356624603, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0171, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.11102861911058426, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0156, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.11057187616825104, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0189, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.11031147837638855, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0155, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.08435497432947159, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0187, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.12450679391622543, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.017, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.09348208457231522, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0171, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.10584745556116104, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0161, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.14194349944591522, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.0177, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.1025288999080658, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0153, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.07800278812646866, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0157, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.1253945231437683, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0176, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.11724173277616501, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.016, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.13043718039989471, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0179, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.08728159964084625, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0177, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.1419975906610489, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0185, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.12774096429347992, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0187, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.10763564705848694, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.0163, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.12671130895614624, + "learning_rate": 1.836929808825424e-05, + "loss": 0.0177, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.10261370241641998, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0145, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12106029689311981, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.0157, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.12428469955921173, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0194, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.08797722309827805, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0175, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.14663904905319214, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.018, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.12288253754377365, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0176, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.11406450718641281, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0164, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10584931075572968, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.017, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1212797686457634, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0176, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.10806597024202347, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0159, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.10098059475421906, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0186, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.14030125737190247, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0204, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.09907040745019913, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0158, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1299905776977539, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0198, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.134843647480011, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0157, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.1048227846622467, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0154, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.12262892723083496, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0193, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.12806116044521332, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0167, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.12875764071941376, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.0149, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.09743807464838028, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0149, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10791302472352982, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0155, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10845673829317093, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0187, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.1366390436887741, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0161, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.101824089884758, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0174, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.12938271462917328, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0159, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.13503864407539368, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0196, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10241002589464188, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0153, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.10051950812339783, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.0159, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11316224932670593, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0166, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.10802372545003891, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0167, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.1365254521369934, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0176, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.11756277829408646, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.016, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13363559544086456, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0167, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.13138234615325928, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0194, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.1519714891910553, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0147, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.12152288109064102, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.017, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.0976492166519165, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0174, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.09356466680765152, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0158, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.09382101893424988, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0175, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.12188217788934708, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0163, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.09245157241821289, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0183, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.1661074161529541, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0202, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.154639333486557, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0143, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.117487333714962, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.017, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.09468945115804672, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0171, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.09632660448551178, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0147, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.12390028685331345, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0177, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.11924377828836441, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0154, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.13191300630569458, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0154, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.07403766363859177, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0171, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.10911701619625092, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0159, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.11888141930103302, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0189, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.09766717255115509, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0182, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.08285287767648697, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0145, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.13635197281837463, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0166, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.11627981066703796, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.10899969190359116, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0174, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.10002870857715607, + "learning_rate": 1.813525107780979e-05, + "loss": 0.016, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09963744133710861, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0163, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.2032729536294937, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0188, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.10394410789012909, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0178, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.1004093736410141, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0181, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.08776404708623886, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.0188, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.10926580429077148, + "learning_rate": 1.811099258182467e-05, + "loss": 0.018, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.09496597945690155, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0153, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.12430521845817566, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0152, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.11681065708398819, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0156, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.10450825840234756, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0162, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10932883620262146, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0163, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.1358717978000641, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0152, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.11744217574596405, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0188, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.14966122806072235, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0179, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.113619364798069, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.08801224827766418, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0168, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.09556800872087479, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0171, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.11116721481084824, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0171, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.10650353878736496, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0158, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.09222083538770676, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0154, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.10958139598369598, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.0189, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11020398885011673, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0166, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.08026755601167679, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0159, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.09629099816083908, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0162, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11310935020446777, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0192, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.1281706690788269, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0191, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12402708828449249, + "learning_rate": 1.802602130181036e-05, + "loss": 0.0159, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.09760920703411102, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0151, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.12342818826436996, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.018, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09423290938138962, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0163, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.10432913154363632, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0172, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.09628765285015106, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0169, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.11590366810560226, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0167, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10148236900568008, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0155, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.10599179565906525, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0155, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.11854606866836548, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.10626410692930222, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0185, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.12172172963619232, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0179, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.10842256247997284, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0173, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.11806859076023102, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0163, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11489225924015045, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0178, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.11859364807605743, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0164, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.18860413134098053, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.0168, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.13375963270664215, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0168, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.13163933157920837, + "learning_rate": 1.795310800633073e-05, + "loss": 0.0151, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.11135797202587128, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0175, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.12392686307430267, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0161, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.0995136946439743, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0165, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.10128366202116013, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0161, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.13080070912837982, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0169, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.10262267291545868, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0176, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.10983351618051529, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0163, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.09834928810596466, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0158, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.11170317232608795, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0166, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.14308896660804749, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.09110505878925323, + "learning_rate": 1.790851399590954e-05, + "loss": 0.017, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.12270113825798035, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0165, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.1422351896762848, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0177, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1294422596693039, + "learning_rate": 1.789634733729274e-05, + "loss": 0.017, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.1157895103096962, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0192, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.12473416328430176, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0156, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.1765025109052658, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0161, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11862069368362427, + "learning_rate": 1.78801220478228e-05, + "loss": 0.017, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.10549142211675644, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0165, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.09818614274263382, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0171, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.13607585430145264, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0168, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.09340739995241165, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0158, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.0805685818195343, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.0207, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.11172603815793991, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0164, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.08757263422012329, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0179, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.09590335935354233, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0199, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.1186867505311966, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.0146, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.09467568248510361, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0175, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11578786373138428, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0179, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09604159742593765, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0173, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.11337848007678986, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.0156, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.11909373849630356, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0188, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.08804072439670563, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0175, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.13454975187778473, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.0184, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.12201394885778427, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0148, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.11946159601211548, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0153, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12322071939706802, + "learning_rate": 1.78030044678245e-05, + "loss": 0.0181, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.10444212704896927, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0159, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.10723353922367096, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0158, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.10750993341207504, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0167, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.1207774356007576, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0207, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.12730205059051514, + "learning_rate": 1.778269750931628e-05, + "loss": 0.017, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.08926276862621307, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0151, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.10322589427232742, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0174, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10034044831991196, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.0184, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.12131599336862564, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0183, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.10715801268815994, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0156, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.14093570411205292, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0169, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.124595507979393, + "learning_rate": 1.775425890028875e-05, + "loss": 0.017, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.13153409957885742, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0189, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.1348494291305542, + "learning_rate": 1.77461316962853e-05, + "loss": 0.019, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.10926292091608047, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0161, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.11094044893980026, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0155, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.1341349333524704, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0184, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.15732111036777496, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0203, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.07597270607948303, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.0184, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.11097310483455658, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.0176, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.12087065726518631, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0164, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12998445332050323, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0191, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.13928557932376862, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0157, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.10487896203994751, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0172, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.13437892496585846, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0182, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.20842674374580383, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0189, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.09813268482685089, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12833304703235626, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0168, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.1019122451543808, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0159, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.07276969403028488, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.016, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09712188690900803, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0169, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.10403666645288467, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0156, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.11877460777759552, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0152, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.10731421411037445, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0188, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.12387207895517349, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0173, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.10561594367027283, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0179, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.2772648334503174, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0168, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.097810760140419, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0142, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.1168239563703537, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0171, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.14195922017097473, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0162, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11385945230722427, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0187, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.1124262586236, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.13074515759944916, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0151, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.12591321766376495, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0177, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.09103676676750183, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0169, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.12457307428121567, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0162, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.11226223409175873, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0156, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.14778819680213928, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0187, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.10433483868837357, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0187, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.08798553794622421, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0163, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.09729046374559402, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0149, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10188018530607224, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0154, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.12559989094734192, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.018, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12567821145057678, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0168, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.12211210280656815, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0166, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.11825426667928696, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0168, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.11676980555057526, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0175, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.11447525024414062, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0162, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.08958559483289719, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0161, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.11528483778238297, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0162, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.08296159654855728, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.014, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.12306056916713715, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0158, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.13315193355083466, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.0164, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.09533339738845825, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0166, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.20049314200878143, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.017, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.14355438947677612, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0187, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.11443490535020828, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0146, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.10336155444383621, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0164, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11926902830600739, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0173, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.09699229151010513, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0176, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.11406200379133224, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0164, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.10817506164312363, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0165, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.11398246139287949, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.0163, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.09427376091480255, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0166, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.10372194647789001, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0148, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.09003781527280807, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0172, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.09807098656892776, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.11895686388015747, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0163, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10839524865150452, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0173, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.20792271196842194, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0183, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.07746267318725586, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0147, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.11470094323158264, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0177, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.1377800703048706, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.1247936338186264, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.016, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.1405080407857895, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0164, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.1318410336971283, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0175, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.14254692196846008, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0159, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.12803269922733307, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0146, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11180103570222855, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0169, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.16513720154762268, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0162, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.13401633501052856, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.0182, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.13510771095752716, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0175, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.12391090393066406, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0174, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1019999086856842, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0157, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.11906205862760544, + "learning_rate": 1.7416306237948e-05, + "loss": 0.0179, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11846558749675751, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.0154, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.12292807549238205, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0163, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10195571929216385, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0186, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.1032365933060646, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0163, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1241045668721199, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0173, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.10956405848264694, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.0163, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.12094716727733612, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.018, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.1353612244129181, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0161, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.0935104638338089, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0164, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.10712502896785736, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.0177, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.12715180218219757, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0148, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.09601753205060959, + "learning_rate": 1.736733684019633e-05, + "loss": 0.0167, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.11189524084329605, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0174, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.11025495082139969, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.016, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.1203494742512703, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0192, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.12150877714157104, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0143, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.11957377195358276, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0146, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.16482087969779968, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0181, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.10647732019424438, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0178, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.13345423340797424, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0155, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.13268540799617767, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0178, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.12209953367710114, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0161, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.11951886862516403, + "learning_rate": 1.73224254699117e-05, + "loss": 0.0213, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.12044327706098557, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0195, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.10611505806446075, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0169, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.12347753345966339, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0145, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13930706679821014, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0185, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.12702012062072754, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0167, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.1193593367934227, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0169, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.16778719425201416, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0179, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11699119210243225, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0159, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11571445316076279, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.08481728285551071, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0159, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.10007977485656738, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0155, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.12851299345493317, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0167, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.13929468393325806, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0154, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.11237664520740509, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0143, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.11654951423406601, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0179, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.11065302044153214, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0168, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.10024461150169373, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0173, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.1334248036146164, + "learning_rate": 1.724888829234145e-05, + "loss": 0.016, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.1115458682179451, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0159, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.15415695309638977, + "learning_rate": 1.72407140459686e-05, + "loss": 0.0166, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10965333878993988, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0159, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.10152038931846619, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0152, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.10030021518468857, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0161, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.1526176780462265, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0134, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13154515624046326, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.017, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.12260743975639343, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0152, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.07395336031913757, + "learning_rate": 1.721209883968656e-05, + "loss": 0.0155, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.1107548177242279, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0178, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.08634423464536667, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0148, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.10544566810131073, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0151, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.13896141946315765, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0171, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10351476818323135, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0164, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.11912661790847778, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0163, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.12101101130247116, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0185, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.09519081562757492, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0137, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.13252407312393188, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0153, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.08964477479457855, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0148, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.11772924661636353, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0153, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1261531263589859, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.0155, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.12295454740524292, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0162, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.14834828674793243, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0162, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.14458344876766205, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0146, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.10006255656480789, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0183, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11527495086193085, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0174, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.11176037788391113, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0146, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.11118462681770325, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0139, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.1216764822602272, + "learning_rate": 1.713029611235338e-05, + "loss": 0.018, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10572952032089233, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0155, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.09299932420253754, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0193, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.11654805392026901, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0146, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.125601127743721, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0148, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.12143341451883316, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0144, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.10944563895463943, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0188, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.10531438142061234, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.017, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.08527344465255737, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0134, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.12489438056945801, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0184, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.12359192967414856, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0162, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.09504237025976181, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0147, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.1144440770149231, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0155, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.11942129582166672, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.0161, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.15496809780597687, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0167, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.11641886830329895, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0155, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.13092041015625, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0164, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.1662251204252243, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0169, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.1151951253414154, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0146, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.13008327782154083, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.017, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.08816810697317123, + "learning_rate": 1.70484286724112e-05, + "loss": 0.0147, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.12602686882019043, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0167, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10360333323478699, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0161, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.10288142412900925, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0174, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.1306314915418625, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0178, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.11859587579965591, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0152, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.12666994333267212, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.0155, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.10859639197587967, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0174, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.1423543244600296, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0145, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.12444380670785904, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0165, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.07491649687290192, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0174, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.12173314392566681, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0156, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.11790800094604492, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.1375807374715805, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0182, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.11894135922193527, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0149, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.09377862513065338, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0157, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1448003053665161, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0189, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.14296171069145203, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.019, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.10862060636281967, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0188, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.1081303209066391, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.0166, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08262617141008377, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0163, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12013579159975052, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0161, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.06014449894428253, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0166, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09688957780599594, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0164, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.11730696260929108, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.0142, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.11243025213479996, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0134, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11974845081567764, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0156, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.1408282220363617, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0193, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.1495579332113266, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.017, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.1408172994852066, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.19624975323677063, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0165, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09940195083618164, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0168, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.11277889460325241, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0171, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.13027498126029968, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0165, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.11570801585912704, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.0159, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.12621626257896423, + "learning_rate": 1.690501241902545e-05, + "loss": 0.016, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.13131381571292877, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0146, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.11577112972736359, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0168, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.1240323930978775, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0162, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.17400938272476196, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.0155, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.09539356827735901, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0165, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10108646005392075, + "learning_rate": 1.688040861067922e-05, + "loss": 0.016, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.1295158863067627, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0151, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.13638637959957123, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0188, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.12255847454071045, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0148, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.13242852687835693, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0166, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11082842200994492, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.015, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.14538930356502533, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0172, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1609591394662857, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0166, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.10611626505851746, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.0142, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.13291358947753906, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0179, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.1112443208694458, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0179, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.10251478105783463, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0174, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.10443990677595139, + "learning_rate": 1.683118563837068e-05, + "loss": 0.016, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.08651736378669739, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0176, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.09996092319488525, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.017, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.13269121944904327, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.0154, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.121453657746315, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0167, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.10587169975042343, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0161, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.12990498542785645, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0158, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.12159401923418045, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0158, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.12046501785516739, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0165, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.10626068711280823, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0144, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.14602169394493103, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0161, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09537681937217712, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0188, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.1014745756983757, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0171, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.11358807235956192, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0165, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.12290225178003311, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0153, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.11178604513406754, + "learning_rate": 1.676962882515376e-05, + "loss": 0.0175, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.1464385688304901, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0155, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.12778854370117188, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0144, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.09556689858436584, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0171, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.10866247862577438, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.0151, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.14120467007160187, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0163, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.12688933312892914, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0174, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.10251957923173904, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0171, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.11465741693973541, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.016, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.08181405812501907, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0201, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.09379345923662186, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0172, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.12122894823551178, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0183, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.0956249088048935, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0142, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.11953603476285934, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0169, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.12494538724422455, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0167, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.20808497071266174, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0138, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.10776039958000183, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0152, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.09606330841779709, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0154, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.10988017171621323, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0182, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.10328707098960876, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0146, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.166867196559906, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0179, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.11701484769582748, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0159, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.10510280728340149, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0179, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.12857040762901306, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0172, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1627577394247055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0176, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.10962127149105072, + "learning_rate": 1.666696746039845e-05, + "loss": 0.0147, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.12961529195308685, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0164, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.08055070787668228, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0148, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.10041354596614838, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0161, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.12196766585111618, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0138, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.0879533439874649, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0156, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.1102190688252449, + "learning_rate": 1.664231677578745e-05, + "loss": 0.016, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.12463485449552536, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.016, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.10829959809780121, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0181, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.08602527529001236, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0176, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.10241527855396271, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.0172, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11007986962795258, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0156, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13110534846782684, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.018, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1417531818151474, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0161, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.13233700394630432, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0147, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.09148544073104858, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0145, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.08781567215919495, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.0149, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.10577751696109772, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0161, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.11689155548810959, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0171, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.0777219608426094, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0152, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.07983718812465668, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0158, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1595761626958847, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0168, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.1037725880742073, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0161, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.10239970684051514, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0149, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.15451422333717346, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0199, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.11882807314395905, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0165, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.09850180149078369, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0147, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.10899021476507187, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.09254994243383408, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0148, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.13559140264987946, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0162, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.11541012674570084, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0154, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.07937203347682953, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0153, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.10062690824270248, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0165, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.13341091573238373, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0178, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.0857294499874115, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0132, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.10551585257053375, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0142, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.12363193184137344, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0165, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.11550596356391907, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0165, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.09379533678293228, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0154, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.10079154372215271, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0163, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.13103018701076508, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0178, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.1172979474067688, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0152, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.11038312315940857, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0152, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.0925338938832283, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0148, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.2519476115703583, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.017, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.09703299403190613, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0193, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.10749004036188126, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0175, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.08736859261989594, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0147, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.13125599920749664, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0161, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10677823424339294, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0156, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.09135954827070236, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0151, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.10548236966133118, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0179, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.1150064468383789, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0188, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.127374529838562, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.016, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.09152884781360626, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0177, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.1021062359213829, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0145, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.1054648831486702, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0172, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.12128820270299911, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0165, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.11375363171100616, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0151, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.11004973948001862, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0182, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.11101444810628891, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0164, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.10393026471138, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0172, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.1139075756072998, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0159, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1569976508617401, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0178, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09649650752544403, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0165, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.16198694705963135, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0157, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.10348767787218094, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0173, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.12259576469659805, + "learning_rate": 1.639145830609088e-05, + "loss": 0.0171, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.10480271279811859, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0159, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.11254257708787918, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0163, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.11534930765628815, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0149, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.10350872576236725, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0169, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.10812029242515564, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.015, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.11796360462903976, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.0144, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.1276526153087616, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0157, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.11665219813585281, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0151, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.10940699279308319, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0165, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.09596745669841766, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0149, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15527229011058807, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0164, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11651212722063065, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0145, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.12004360556602478, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0153, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.13298730552196503, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0176, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.1496417224407196, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0153, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.11104541271924973, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0158, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.12215209007263184, + "learning_rate": 1.632147383062092e-05, + "loss": 0.015, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.09335075318813324, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0148, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.11481857299804688, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0153, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14369069039821625, + "learning_rate": 1.630912057457699e-05, + "loss": 0.019, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.11009083688259125, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0162, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.11112431436777115, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.0137, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.10135089606046677, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0167, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09966760873794556, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0168, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12062226980924606, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0145, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.1109282448887825, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0162, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.10862662643194199, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0152, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.09910363703966141, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0156, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.1374131590127945, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0167, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.12114953994750977, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.0156, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.14391179382801056, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0156, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.12364984303712845, + "learning_rate": 1.625969868712804e-05, + "loss": 0.0169, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.08771048486232758, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0142, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.10959397256374359, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0176, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.138074591755867, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0149, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.1116161122918129, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0154, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.1398090422153473, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0164, + "step": 60000 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 20000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 64, + "trial_name": null, + "trial_params": null +} diff --git a/checkpoint-80000/config.json b/checkpoint-80000/config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d7b77e4255263f3803a4b6f94ca6831e459b70 --- /dev/null +++ b/checkpoint-80000/config.json @@ -0,0 +1,64 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/checkpoint-80000/experiment_cfg/metadata.json b/checkpoint-80000/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..300eb070963337016e16465f50ba8351ea2eb25c --- /dev/null +++ b/checkpoint-80000/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.3139495849609375, + 0.4876587688922882, + 0.7196521759033203 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6996827125549316 + ], + "mean": [ + 2.749288365724937, + -2.0455216239909983, + 0.700532551020533 + ], + "std": [ + 1.6786295897046262, + 1.312897969434244, + 0.00133751758906258 + ], + "q01": [ + -1.4241907881148037, + -5.1716547935950885, + 0.6999670898768614 + ], + "q99": [ + 6.198111545831555, + -0.5873531334104489, + 0.7038833014242587 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.23084999587450247, + 0.6238431088581847 + ], + "std": [ + 0.0, + 0.0, + 0.6556404371644047, + 0.3572252105732042 + ], + "q01": [ + 0.0, + 0.0, + -0.9999999999999999, + 1.7482354593273349e-06 + ], + "q99": [ + 0.0, + 0.0, + 0.9999999999999999, + 0.9999999999999999 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8003877401351929, + 0.9829942584037781 + ], + "min": [ + -0.37471598386764526, + -0.8472502827644348, + -0.25070279836654663 + ], + "mean": [ + 0.28667332601863704, + -0.038315473170876704, + 0.45974658906777444 + ], + "std": [ + 0.17067058828305154, + 0.23569952037784195, + 0.21950702354181487 + ], + "q01": [ + 0.015172948963784228, + -0.44450502063220615, + 0.23314444103329338 + ], + "q99": [ + 0.5609069689572412, + 0.38454179128009547, + 0.7028102736473852 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9984455108642578, + 0.9434727430343628, + 0.9062229990959167 + ], + "min": [ + -0.9999930262565613, + -0.9988337755203247, + -0.9618149995803833, + 2.1872274658107926e-07 + ], + "mean": [ + -0.25672297401336, + 0.024425335806687216, + -0.08740977164483847, + 0.16608739644018397 + ], + "std": [ + 0.7932585208392383, + 0.3102260195580416, + 0.3772472576315148, + 0.17451959300158773 + ], + "q01": [ + -0.99379006987882, + -0.5478102050038828, + -0.6101777952971636, + 0.002274085014134748 + ], + "q99": [ + 0.8804856038896288, + 0.5910611092873037, + 0.5216058042993562, + 0.5231965974012255 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011436971835792065, + -0.05664053186774254 + ], + "mean": [ + 0.031651551516052555, + -0.03161481387920421 + ], + "std": [ + 0.013119769222217526, + 0.01306078225192402 + ], + "q01": [ + 0.0060999246446777336, + -0.04062601653964198 + ], + "q99": [ + 0.04054891621517283, + -0.0059163567113456345 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.008144896753718373, + -0.00018893135146078263, + -0.0008739062727221845, + 0.0 + ], + "std": [ + 0.11030045424364411, + 0.10148082570313594, + 0.08975698373467506, + 0.0 + ], + "q01": [ + -0.07287373067581518, + -0.09446994199569948, + -0.07702818259249572, + 0.0 + ], + "q99": [ + 0.04485485414407898, + 0.09626941605965177, + 0.07610665906122742, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9221974313425939 + ], + "std": [ + 0.38670194342612674 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.384693946883099 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.01976129923017491, + -0.020870631344490034, + -0.05993476517349181 + ], + "std": [ + 0.4347024990804053, + 0.4221662976569997, + 0.3845506843044066 + ], + "q01": [ + -0.8835792939767807, + -0.9280919251713778, + -0.783361071470956 + ], + "q99": [ + 0.7622084438449307, + 0.8625125252609077, + 0.7470974753250706 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.008089673068544335, + -0.026498254627927348, + 0.0029083551743059005 + ], + "std": [ + 0.11216734503733773, + 0.13206002613798629, + 0.12615870412692864 + ], + "q01": [ + -0.2814627972846672, + -0.4342386516115439, + -0.34489755628340574 + ], + "q99": [ + 0.31879353266939386, + 0.28411797683220563, + 0.3597718671669894 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.3824228874769045 + ], + "std": [ + 0.9240015096469786 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.5597272305813592 + ] + } + } + }, + "modalities": { + "video": { + "robot0_eye_in_hand": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_left": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_right": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": true, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + }, + "end_effector_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": true, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/checkpoint-80000/model-00001-of-00002.safetensors b/checkpoint-80000/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0f9df061051b36dc6222fd8f052d5b4faf8ce7cb --- /dev/null +++ b/checkpoint-80000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7926db655b83d22f93f261f84206dc15e9cbb0c829de9ce2f4d99db256e39c09 +size 4999367032 diff --git a/checkpoint-80000/model-00002-of-00002.safetensors b/checkpoint-80000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..902778266b7ba23bc447f12c0cf152b2b11cf526 --- /dev/null +++ b/checkpoint-80000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c2421c6b0e913b02c3c31dea19624ccc77a54004f8a9f1f2cf630af360abec6a +size 2586705312 diff --git a/checkpoint-80000/model.safetensors.index.json b/checkpoint-80000/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/checkpoint-80000/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/checkpoint-80000/optimizer.pt b/checkpoint-80000/optimizer.pt new file mode 100644 index 0000000000000000000000000000000000000000..e57aff246457948aa9712ca9d383941497bd2467 --- /dev/null +++ b/checkpoint-80000/optimizer.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8173160b8876a0409960c5fb85dd2f1249fdbe20b1d37d8128207a7901fd9af +size 8550719959 diff --git a/checkpoint-80000/rng_state_0.pth b/checkpoint-80000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..fe049471e4fcff2ae65c50b665f95fe7f02898eb --- /dev/null +++ b/checkpoint-80000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:225e23d5e4b7ca5739ae397074840e33a5b82ae6797ef28d800e15616e9c3dc1 +size 14917 diff --git a/checkpoint-80000/rng_state_1.pth b/checkpoint-80000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..ed770f3ac1909abf22c43ae4eb21af5e5937362e --- /dev/null +++ b/checkpoint-80000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:734e952eba5e2758d6b2fc3b74248c3e747d7fd747bd03e608b6e5932e6bcdbb +size 14917 diff --git a/checkpoint-80000/scheduler.pt b/checkpoint-80000/scheduler.pt new file mode 100644 index 0000000000000000000000000000000000000000..5f2e8444b7dcc93a2a89577b2f73c67eb6d4d74d --- /dev/null +++ b/checkpoint-80000/scheduler.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:46663a5d5431f344b06fa59c078feb4aa842edc31ef2bd2d5fc3ad70015bffea +size 1465 diff --git a/checkpoint-80000/trainer_state.json b/checkpoint-80000/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..bed7f2cc864bd3dca729d990b7aa0f12150cbc7e --- /dev/null +++ b/checkpoint-80000/trainer_state.json @@ -0,0 +1,56034 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 23.276112889147512, + "eval_steps": 500, + "global_step": 80000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002909514111143439, + "grad_norm": 5.5908684730529785, + "learning_rate": 4.5e-08, + "loss": 0.5668, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.084368705749512, + "learning_rate": 9.5e-08, + "loss": 0.5781, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 5.755157947540283, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5759, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.174816608428955, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5202, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 6.199656963348389, + "learning_rate": 2.45e-07, + "loss": 0.5455, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.251872539520264, + "learning_rate": 2.95e-07, + "loss": 0.5507, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 5.606583595275879, + "learning_rate": 3.45e-07, + "loss": 0.5264, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.967868328094482, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.5166, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.92909574508667, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4578, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.306433200836182, + "learning_rate": 4.95e-07, + "loss": 0.4091, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.75846529006958, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4072, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.2030856609344482, + "learning_rate": 5.95e-07, + "loss": 0.2992, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.5756216049194336, + "learning_rate": 6.45e-07, + "loss": 0.2703, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.3420923948287964, + "learning_rate": 6.95e-07, + "loss": 0.2492, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 0.9191234111785889, + "learning_rate": 7.45e-07, + "loss": 0.2302, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.662022054195404, + "learning_rate": 7.95e-07, + "loss": 0.1993, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.704409658908844, + "learning_rate": 8.45e-07, + "loss": 0.1814, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.6104884147644043, + "learning_rate": 8.95e-07, + "loss": 0.1694, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.4955693781375885, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1674, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.707667887210846, + "learning_rate": 9.95e-07, + "loss": 0.1507, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.6977017521858215, + "learning_rate": 1.045e-06, + "loss": 0.1547, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.49085867404937744, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1536, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.6363996267318726, + "learning_rate": 1.145e-06, + "loss": 0.1565, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.4633055627346039, + "learning_rate": 1.195e-06, + "loss": 0.1641, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5155240297317505, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1513, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5815547108650208, + "learning_rate": 1.295e-06, + "loss": 0.1431, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.5865835547447205, + "learning_rate": 1.345e-06, + "loss": 0.1419, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.5872474312782288, + "learning_rate": 1.395e-06, + "loss": 0.1326, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.4897519052028656, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1326, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.46316635608673096, + "learning_rate": 1.495e-06, + "loss": 0.1442, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.5359047651290894, + "learning_rate": 1.545e-06, + "loss": 0.1394, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.5805708169937134, + "learning_rate": 1.595e-06, + "loss": 0.1231, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.5488415360450745, + "learning_rate": 1.645e-06, + "loss": 0.127, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.41755908727645874, + "learning_rate": 1.695e-06, + "loss": 0.1375, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.6258692145347595, + "learning_rate": 1.745e-06, + "loss": 0.1202, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.5017617344856262, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1317, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.710078239440918, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.46977895498275757, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1215, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.4802077114582062, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.116, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.6448793411254883, + "learning_rate": 1.995e-06, + "loss": 0.1172, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4394509494304657, + "learning_rate": 2.045e-06, + "loss": 0.1145, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.6547806859016418, + "learning_rate": 2.095e-06, + "loss": 0.1174, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.6954914331436157, + "learning_rate": 2.145e-06, + "loss": 0.1245, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.670100212097168, + "learning_rate": 2.195e-06, + "loss": 0.1133, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.5230806469917297, + "learning_rate": 2.245e-06, + "loss": 0.11, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.46809589862823486, + "learning_rate": 2.295e-06, + "loss": 0.1069, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.6178835034370422, + "learning_rate": 2.345e-06, + "loss": 0.1099, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.41567298769950867, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1072, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4968737065792084, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1093, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.46893924474716187, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1046, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.5041665434837341, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1014, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5833673477172852, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1095, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.5565881133079529, + "learning_rate": 2.645e-06, + "loss": 0.1057, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6213339567184448, + "learning_rate": 2.695e-06, + "loss": 0.1087, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.5638794898986816, + "learning_rate": 2.745e-06, + "loss": 0.099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.5594047904014587, + "learning_rate": 2.795e-06, + "loss": 0.1065, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5521039366722107, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1072, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.5013270378112793, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1045, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.44355231523513794, + "learning_rate": 2.945e-06, + "loss": 0.0985, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.5467342734336853, + "learning_rate": 2.995e-06, + "loss": 0.101, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5065611600875854, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0939, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.5587425827980042, + "learning_rate": 3.095e-06, + "loss": 0.0969, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5561433434486389, + "learning_rate": 3.145e-06, + "loss": 0.1009, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.5576680302619934, + "learning_rate": 3.195e-06, + "loss": 0.1006, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.7426344752311707, + "learning_rate": 3.245e-06, + "loss": 0.0959, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.64913409948349, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.0962, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.5271337628364563, + "learning_rate": 3.345e-06, + "loss": 0.0927, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.6992810368537903, + "learning_rate": 3.395e-06, + "loss": 0.0924, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.559481680393219, + "learning_rate": 3.445e-06, + "loss": 0.0886, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7303935885429382, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0918, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.45654281973838806, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0934, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.41098082065582275, + "learning_rate": 3.595e-06, + "loss": 0.0911, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.43506675958633423, + "learning_rate": 3.645e-06, + "loss": 0.0887, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5131197571754456, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0983, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.4571124017238617, + "learning_rate": 3.745e-06, + "loss": 0.0928, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.6749465465545654, + "learning_rate": 3.795e-06, + "loss": 0.0926, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.46945369243621826, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1016, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.6481541395187378, + "learning_rate": 3.895e-06, + "loss": 0.0917, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.6769195795059204, + "learning_rate": 3.945e-06, + "loss": 0.0924, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5225585103034973, + "learning_rate": 3.995e-06, + "loss": 0.0909, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.642867922782898, + "learning_rate": 4.045e-06, + "loss": 0.0829, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.4617709219455719, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0924, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.5350925326347351, + "learning_rate": 4.145e-06, + "loss": 0.0821, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.4871961176395416, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0896, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.5545480847358704, + "learning_rate": 4.245e-06, + "loss": 0.0891, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.6057930588722229, + "learning_rate": 4.295e-06, + "loss": 0.0876, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.515282154083252, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0873, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.5402254462242126, + "learning_rate": 4.395e-06, + "loss": 0.0908, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4669134318828583, + "learning_rate": 4.445e-06, + "loss": 0.0837, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.46819189190864563, + "learning_rate": 4.495e-06, + "loss": 0.0888, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35766035318374634, + "learning_rate": 4.545e-06, + "loss": 0.0829, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.6769008040428162, + "learning_rate": 4.595e-06, + "loss": 0.0817, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5884619355201721, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0892, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5613101720809937, + "learning_rate": 4.695e-06, + "loss": 0.0852, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.47087088227272034, + "learning_rate": 4.745e-06, + "loss": 0.0818, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.4887458384037018, + "learning_rate": 4.795e-06, + "loss": 0.0858, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.44256526231765747, + "learning_rate": 4.845e-06, + "loss": 0.0833, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5375688076019287, + "learning_rate": 4.895e-06, + "loss": 0.075, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.6297019124031067, + "learning_rate": 4.945e-06, + "loss": 0.0932, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.6576125621795654, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0829, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.724773108959198, + "learning_rate": 5.045e-06, + "loss": 0.0813, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.54575115442276, + "learning_rate": 5.095e-06, + "loss": 0.0873, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.5186223387718201, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0736, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.46651381254196167, + "learning_rate": 5.195e-06, + "loss": 0.0791, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.5323564410209656, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0753, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.42579638957977295, + "learning_rate": 5.295e-06, + "loss": 0.0755, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.5063076019287109, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0767, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.4558819532394409, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0807, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.6920697093009949, + "learning_rate": 5.445e-06, + "loss": 0.079, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.4706108868122101, + "learning_rate": 5.495e-06, + "loss": 0.079, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.46705493330955505, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0745, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4643767774105072, + "learning_rate": 5.595e-06, + "loss": 0.0778, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.4800759553909302, + "learning_rate": 5.645e-06, + "loss": 0.0792, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.37705323100090027, + "learning_rate": 5.695e-06, + "loss": 0.0815, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.6651142239570618, + "learning_rate": 5.745e-06, + "loss": 0.0732, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.5275036692619324, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0758, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.45047008991241455, + "learning_rate": 5.845e-06, + "loss": 0.0808, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.6931971311569214, + "learning_rate": 5.895e-06, + "loss": 0.0808, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.7053916454315186, + "learning_rate": 5.945e-06, + "loss": 0.0747, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.6798260807991028, + "learning_rate": 5.995e-06, + "loss": 0.0829, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.5724003314971924, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0749, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.5836488008499146, + "learning_rate": 6.095e-06, + "loss": 0.0759, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.40817850828170776, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0778, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.41070470213890076, + "learning_rate": 6.195e-06, + "loss": 0.0765, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.43642961978912354, + "learning_rate": 6.245e-06, + "loss": 0.0711, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.49526578187942505, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0713, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.7773705720901489, + "learning_rate": 6.345e-06, + "loss": 0.0744, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.5485936999320984, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0732, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.5593521595001221, + "learning_rate": 6.445e-06, + "loss": 0.0747, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.3667270839214325, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0723, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.5555917620658875, + "learning_rate": 6.545e-06, + "loss": 0.0726, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.38107141852378845, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0647, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.46452224254608154, + "learning_rate": 6.645e-06, + "loss": 0.0787, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6018183827400208, + "learning_rate": 6.695e-06, + "loss": 0.0718, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.44997578859329224, + "learning_rate": 6.745e-06, + "loss": 0.0758, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45717573165893555, + "learning_rate": 6.795e-06, + "loss": 0.0738, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.5185182690620422, + "learning_rate": 6.845e-06, + "loss": 0.0768, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.5135819315910339, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.43529245257377625, + "learning_rate": 6.945e-06, + "loss": 0.0691, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.49389752745628357, + "learning_rate": 6.995e-06, + "loss": 0.0678, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.4113156199455261, + "learning_rate": 7.045e-06, + "loss": 0.0651, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.4072611927986145, + "learning_rate": 7.095e-06, + "loss": 0.0691, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.4524228870868683, + "learning_rate": 7.145e-06, + "loss": 0.0688, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.6225605607032776, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0676, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4841763377189636, + "learning_rate": 7.245e-06, + "loss": 0.0678, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.46297937631607056, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0759, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.5585324168205261, + "learning_rate": 7.345e-06, + "loss": 0.0671, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.57328200340271, + "learning_rate": 7.395e-06, + "loss": 0.0717, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.5072467923164368, + "learning_rate": 7.445000000000001e-06, + "loss": 0.071, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.4630240797996521, + "learning_rate": 7.495e-06, + "loss": 0.0674, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.35908272862434387, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0704, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.5117342472076416, + "learning_rate": 7.595e-06, + "loss": 0.0675, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.5214630961418152, + "learning_rate": 7.645e-06, + "loss": 0.0702, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.39036089181900024, + "learning_rate": 7.695e-06, + "loss": 0.0709, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.558722734451294, + "learning_rate": 7.745e-06, + "loss": 0.0705, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.3612176179885864, + "learning_rate": 7.795e-06, + "loss": 0.0715, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.523444652557373, + "learning_rate": 7.845e-06, + "loss": 0.0685, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.5231291651725769, + "learning_rate": 7.895e-06, + "loss": 0.0717, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.50368332862854, + "learning_rate": 7.945e-06, + "loss": 0.0694, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.4126806855201721, + "learning_rate": 7.995e-06, + "loss": 0.0693, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.719352662563324, + "learning_rate": 8.045e-06, + "loss": 0.0651, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.4405677020549774, + "learning_rate": 8.095e-06, + "loss": 0.0645, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.34746286273002625, + "learning_rate": 8.145e-06, + "loss": 0.0687, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.43200597167015076, + "learning_rate": 8.195e-06, + "loss": 0.0655, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.3464098274707794, + "learning_rate": 8.245e-06, + "loss": 0.0639, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.27263233065605164, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0629, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.343181848526001, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3949901759624481, + "learning_rate": 8.395e-06, + "loss": 0.0643, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.4489574730396271, + "learning_rate": 8.445e-06, + "loss": 0.0698, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.4746992588043213, + "learning_rate": 8.495e-06, + "loss": 0.0702, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.3651655316352844, + "learning_rate": 8.545e-06, + "loss": 0.0673, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.471071720123291, + "learning_rate": 8.595e-06, + "loss": 0.0648, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.35623008012771606, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0698, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.41837432980537415, + "learning_rate": 8.695e-06, + "loss": 0.0639, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.45608940720558167, + "learning_rate": 8.745e-06, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.40346869826316833, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0625, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.479824960231781, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0703, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.38570791482925415, + "learning_rate": 8.895e-06, + "loss": 0.0643, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.36530473828315735, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0652, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37868866324424744, + "learning_rate": 8.995000000000001e-06, + "loss": 0.068, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.32546791434288025, + "learning_rate": 9.045e-06, + "loss": 0.0636, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4664970338344574, + "learning_rate": 9.095e-06, + "loss": 0.0671, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4767569601535797, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0625, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.41823142766952515, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0685, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.3709985613822937, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0681, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.3933296799659729, + "learning_rate": 9.295e-06, + "loss": 0.0678, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.3079732060432434, + "learning_rate": 9.345e-06, + "loss": 0.0598, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.33485978841781616, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0644, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.40312913060188293, + "learning_rate": 9.445e-06, + "loss": 0.0684, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.4334392547607422, + "learning_rate": 9.495e-06, + "loss": 0.0622, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.5787541270256042, + "learning_rate": 9.545e-06, + "loss": 0.0656, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.2827294170856476, + "learning_rate": 9.595e-06, + "loss": 0.0628, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4431678056716919, + "learning_rate": 9.645e-06, + "loss": 0.0633, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.3916102647781372, + "learning_rate": 9.695e-06, + "loss": 0.0632, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.41470348834991455, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0612, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.3224067986011505, + "learning_rate": 9.795e-06, + "loss": 0.061, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.4110186696052551, + "learning_rate": 9.845e-06, + "loss": 0.066, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.4786360561847687, + "learning_rate": 9.895e-06, + "loss": 0.0648, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3667179048061371, + "learning_rate": 9.945e-06, + "loss": 0.0663, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.31722185015678406, + "learning_rate": 9.995e-06, + "loss": 0.0588, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.29271745681762695, + "learning_rate": 1.0045e-05, + "loss": 0.0636, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.3536592423915863, + "learning_rate": 1.0095e-05, + "loss": 0.0625, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.3228510916233063, + "learning_rate": 1.0145e-05, + "loss": 0.067, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.38474392890930176, + "learning_rate": 1.0195e-05, + "loss": 0.0609, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.40410760045051575, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0638, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.3632320463657379, + "learning_rate": 1.0295e-05, + "loss": 0.0587, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.37775692343711853, + "learning_rate": 1.0345e-05, + "loss": 0.0673, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.3343006372451782, + "learning_rate": 1.0395e-05, + "loss": 0.0599, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3232511281967163, + "learning_rate": 1.0445e-05, + "loss": 0.0661, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.38730430603027344, + "learning_rate": 1.0495e-05, + "loss": 0.0649, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.38972100615501404, + "learning_rate": 1.0545e-05, + "loss": 0.0582, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.27242428064346313, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0583, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.30514073371887207, + "learning_rate": 1.0645e-05, + "loss": 0.065, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.45364341139793396, + "learning_rate": 1.0695e-05, + "loss": 0.0576, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.3863564729690552, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0579, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.33717092871665955, + "learning_rate": 1.0795e-05, + "loss": 0.0565, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38073790073394775, + "learning_rate": 1.0845e-05, + "loss": 0.0585, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4277617037296295, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0622, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.3545698821544647, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0609, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.30646640062332153, + "learning_rate": 1.0995e-05, + "loss": 0.0659, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.3607352375984192, + "learning_rate": 1.1045e-05, + "loss": 0.0621, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3624425530433655, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0626, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3856692612171173, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0604, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.4390200972557068, + "learning_rate": 1.1195e-05, + "loss": 0.0601, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.4108872711658478, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0618, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.32732832431793213, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0636, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.3602239787578583, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.29978248476982117, + "learning_rate": 1.1395e-05, + "loss": 0.058, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.48490920662879944, + "learning_rate": 1.1445e-05, + "loss": 0.0614, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.3278113901615143, + "learning_rate": 1.1495e-05, + "loss": 0.0611, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.37862712144851685, + "learning_rate": 1.1545e-05, + "loss": 0.0626, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.3514714241027832, + "learning_rate": 1.1595e-05, + "loss": 0.0583, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.27256545424461365, + "learning_rate": 1.1645e-05, + "loss": 0.062, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.3305833339691162, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0572, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.2881031930446625, + "learning_rate": 1.1745e-05, + "loss": 0.0641, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.3219377398490906, + "learning_rate": 1.1795e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.34189504384994507, + "learning_rate": 1.1845e-05, + "loss": 0.0607, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30541297793388367, + "learning_rate": 1.1895e-05, + "loss": 0.056, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.35898879170417786, + "learning_rate": 1.1945e-05, + "loss": 0.0598, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.33221277594566345, + "learning_rate": 1.1995e-05, + "loss": 0.0603, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.3435590863227844, + "learning_rate": 1.2045e-05, + "loss": 0.0553, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.24820426106452942, + "learning_rate": 1.2095e-05, + "loss": 0.0644, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.3749466836452484, + "learning_rate": 1.2145e-05, + "loss": 0.0615, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.3984697759151459, + "learning_rate": 1.2195e-05, + "loss": 0.0595, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3250048756599426, + "learning_rate": 1.2245e-05, + "loss": 0.0608, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.2794302701950073, + "learning_rate": 1.2295e-05, + "loss": 0.0578, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3811834752559662, + "learning_rate": 1.2345e-05, + "loss": 0.0612, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3160156309604645, + "learning_rate": 1.2395e-05, + "loss": 0.0541, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.3825189173221588, + "learning_rate": 1.2445e-05, + "loss": 0.0621, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.3966415226459503, + "learning_rate": 1.2495e-05, + "loss": 0.0598, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3615199625492096, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0616, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.440762996673584, + "learning_rate": 1.2595e-05, + "loss": 0.0599, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.345599889755249, + "learning_rate": 1.2645e-05, + "loss": 0.0597, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.2963278889656067, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0556, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.3720734417438507, + "learning_rate": 1.2745e-05, + "loss": 0.0618, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3937445878982544, + "learning_rate": 1.2795e-05, + "loss": 0.0624, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.4492146372795105, + "learning_rate": 1.2845e-05, + "loss": 0.0573, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.5172158479690552, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0594, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.3796320855617523, + "learning_rate": 1.2945e-05, + "loss": 0.0528, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.42423033714294434, + "learning_rate": 1.2995e-05, + "loss": 0.0615, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.4578545093536377, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0562, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.38345491886138916, + "learning_rate": 1.3095e-05, + "loss": 0.0574, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.2539816200733185, + "learning_rate": 1.3145e-05, + "loss": 0.0562, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.44323891401290894, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0588, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.4074377417564392, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0591, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.5252887606620789, + "learning_rate": 1.3295e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.39536044001579285, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0588, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.6300645470619202, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0584, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.312466561794281, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0588, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3810383379459381, + "learning_rate": 1.3495e-05, + "loss": 0.0597, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.30645400285720825, + "learning_rate": 1.3545e-05, + "loss": 0.0589, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.2857605218887329, + "learning_rate": 1.3595e-05, + "loss": 0.0589, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.3206622898578644, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0597, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3226073086261749, + "learning_rate": 1.3695e-05, + "loss": 0.0559, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.24912619590759277, + "learning_rate": 1.3745e-05, + "loss": 0.0547, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.29700762033462524, + "learning_rate": 1.3795e-05, + "loss": 0.0563, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.32627588510513306, + "learning_rate": 1.3845e-05, + "loss": 0.0602, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.48569780588150024, + "learning_rate": 1.3895e-05, + "loss": 0.0553, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.323607474565506, + "learning_rate": 1.3945e-05, + "loss": 0.0578, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.42626142501831055, + "learning_rate": 1.3995e-05, + "loss": 0.0539, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.3281888961791992, + "learning_rate": 1.4045e-05, + "loss": 0.0638, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.38216397166252136, + "learning_rate": 1.4095e-05, + "loss": 0.0572, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.2816051244735718, + "learning_rate": 1.4145e-05, + "loss": 0.0543, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.2604160010814667, + "learning_rate": 1.4195e-05, + "loss": 0.0555, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.34318771958351135, + "learning_rate": 1.4245e-05, + "loss": 0.0565, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.31248190999031067, + "learning_rate": 1.4295e-05, + "loss": 0.0583, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.3014426827430725, + "learning_rate": 1.4345e-05, + "loss": 0.0586, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3650360703468323, + "learning_rate": 1.4395e-05, + "loss": 0.0581, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.33519116044044495, + "learning_rate": 1.4445e-05, + "loss": 0.0611, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.3595008850097656, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.054, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.32799217104911804, + "learning_rate": 1.4545e-05, + "loss": 0.0574, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.29410457611083984, + "learning_rate": 1.4595e-05, + "loss": 0.0582, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.2727576792240143, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0588, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.36562401056289673, + "learning_rate": 1.4695e-05, + "loss": 0.0597, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.3351168930530548, + "learning_rate": 1.4745e-05, + "loss": 0.0558, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.23880337178707123, + "learning_rate": 1.4795e-05, + "loss": 0.0571, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.3910059928894043, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0573, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.3103587031364441, + "learning_rate": 1.4895e-05, + "loss": 0.0578, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.3265620768070221, + "learning_rate": 1.4945e-05, + "loss": 0.0542, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.3060731291770935, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0537, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.41589322686195374, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0546, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.3102424740791321, + "learning_rate": 1.5095e-05, + "loss": 0.0564, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.32286810874938965, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0519, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.3129369020462036, + "learning_rate": 1.5195e-05, + "loss": 0.0566, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.43735018372535706, + "learning_rate": 1.5245e-05, + "loss": 0.0552, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.440449595451355, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0589, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.3631616532802582, + "learning_rate": 1.5345e-05, + "loss": 0.0511, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.29124221205711365, + "learning_rate": 1.5395e-05, + "loss": 0.055, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.4247622489929199, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0562, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.36976534128189087, + "learning_rate": 1.5495e-05, + "loss": 0.0604, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.3580514192581177, + "learning_rate": 1.5545e-05, + "loss": 0.0547, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.2969493567943573, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0564, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28583595156669617, + "learning_rate": 1.5645e-05, + "loss": 0.0586, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.3220328390598297, + "learning_rate": 1.5695e-05, + "loss": 0.0588, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.32270652055740356, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0559, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.3170712888240814, + "learning_rate": 1.5795e-05, + "loss": 0.0566, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.2966325879096985, + "learning_rate": 1.5845e-05, + "loss": 0.0551, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.2788766622543335, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.051, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.3521708548069, + "learning_rate": 1.5945e-05, + "loss": 0.0558, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.35105040669441223, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0551, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.4221852123737335, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0573, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.4217511713504791, + "learning_rate": 1.6095e-05, + "loss": 0.0532, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.4519175589084625, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0489, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3427059054374695, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0544, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29204127192497253, + "learning_rate": 1.6245e-05, + "loss": 0.0525, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.24138635396957397, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0559, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.3599991500377655, + "learning_rate": 1.6345e-05, + "loss": 0.0544, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.3326258361339569, + "learning_rate": 1.6395e-05, + "loss": 0.0593, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31976261734962463, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0624, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.26379555463790894, + "learning_rate": 1.6495e-05, + "loss": 0.0529, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2576923966407776, + "learning_rate": 1.6545e-05, + "loss": 0.0527, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.30890268087387085, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0571, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.4547998309135437, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.3050350248813629, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0552, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.2984176278114319, + "learning_rate": 1.6745e-05, + "loss": 0.0528, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.34583792090415955, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0512, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.33085328340530396, + "learning_rate": 1.6845e-05, + "loss": 0.0563, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.33391326665878296, + "learning_rate": 1.6895e-05, + "loss": 0.0527, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.23657914996147156, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0509, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.3719659149646759, + "learning_rate": 1.6995e-05, + "loss": 0.0568, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.4268389940261841, + "learning_rate": 1.7045e-05, + "loss": 0.0525, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.2726958394050598, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0482, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27847009897232056, + "learning_rate": 1.7145e-05, + "loss": 0.0559, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.7195e-05, + "loss": 0.0588, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.2666028141975403, + "learning_rate": 1.7245e-05, + "loss": 0.0548, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.26048436760902405, + "learning_rate": 1.7295e-05, + "loss": 0.0592, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.22569318115711212, + "learning_rate": 1.7345e-05, + "loss": 0.0505, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.35350096225738525, + "learning_rate": 1.7395e-05, + "loss": 0.0566, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.3166314363479614, + "learning_rate": 1.7445e-05, + "loss": 0.0517, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.24918541312217712, + "learning_rate": 1.7495e-05, + "loss": 0.0545, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3416207730770111, + "learning_rate": 1.7545e-05, + "loss": 0.0544, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.22633257508277893, + "learning_rate": 1.7595e-05, + "loss": 0.0545, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.3972736597061157, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0591, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3213656544685364, + "learning_rate": 1.7695e-05, + "loss": 0.0537, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.30361926555633545, + "learning_rate": 1.7745e-05, + "loss": 0.0529, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.35521000623703003, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0549, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.4444983899593353, + "learning_rate": 1.7845e-05, + "loss": 0.0529, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.32495826482772827, + "learning_rate": 1.7895e-05, + "loss": 0.0557, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.2893896698951721, + "learning_rate": 1.7945e-05, + "loss": 0.056, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.2607886493206024, + "learning_rate": 1.7995e-05, + "loss": 0.055, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.3305937647819519, + "learning_rate": 1.8045e-05, + "loss": 0.0529, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.3003319799900055, + "learning_rate": 1.8095e-05, + "loss": 0.0515, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3523712456226349, + "learning_rate": 1.8145e-05, + "loss": 0.0524, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.27575406432151794, + "learning_rate": 1.8195e-05, + "loss": 0.0535, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.36445286870002747, + "learning_rate": 1.8245e-05, + "loss": 0.0552, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.41415491700172424, + "learning_rate": 1.8295e-05, + "loss": 0.0537, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.3738953471183777, + "learning_rate": 1.8345e-05, + "loss": 0.0555, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.3446105420589447, + "learning_rate": 1.8395e-05, + "loss": 0.0526, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.39016082882881165, + "learning_rate": 1.8445e-05, + "loss": 0.0536, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.4075721502304077, + "learning_rate": 1.8495e-05, + "loss": 0.0548, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.4071345925331116, + "learning_rate": 1.8545e-05, + "loss": 0.0539, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.34446755051612854, + "learning_rate": 1.8595e-05, + "loss": 0.057, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.3431609272956848, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0515, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.31537994742393494, + "learning_rate": 1.8695e-05, + "loss": 0.0534, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.4846970736980438, + "learning_rate": 1.8745e-05, + "loss": 0.0532, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.3109683394432068, + "learning_rate": 1.8795e-05, + "loss": 0.0547, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.28903284668922424, + "learning_rate": 1.8845e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.26592350006103516, + "learning_rate": 1.8895e-05, + "loss": 0.0499, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.27347302436828613, + "learning_rate": 1.8945e-05, + "loss": 0.052, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.3194405138492584, + "learning_rate": 1.8995e-05, + "loss": 0.0521, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.40457430481910706, + "learning_rate": 1.9045e-05, + "loss": 0.0523, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.3055919408798218, + "learning_rate": 1.9095e-05, + "loss": 0.0504, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.4911873936653137, + "learning_rate": 1.9145e-05, + "loss": 0.0531, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.30619072914123535, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0525, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.24404728412628174, + "learning_rate": 1.9245e-05, + "loss": 0.0498, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.31723177433013916, + "learning_rate": 1.9295e-05, + "loss": 0.0519, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.28105342388153076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0528, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.32965341210365295, + "learning_rate": 1.9395e-05, + "loss": 0.0504, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.2487417608499527, + "learning_rate": 1.9445e-05, + "loss": 0.0552, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3788810670375824, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0535, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.2561226189136505, + "learning_rate": 1.9545e-05, + "loss": 0.056, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.29964715242385864, + "learning_rate": 1.9595e-05, + "loss": 0.0565, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.386600136756897, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0468, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.37690475583076477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.35367539525032043, + "learning_rate": 1.9745e-05, + "loss": 0.0501, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.30189844965934753, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0524, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.27525416016578674, + "learning_rate": 1.9845e-05, + "loss": 0.0524, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.22557462751865387, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.3572654128074646, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0536, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.37130483984947205, + "learning_rate": 1.9995e-05, + "loss": 0.0507, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.40820422768592834, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0547, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.24222403764724731, + "learning_rate": 2.0095e-05, + "loss": 0.0536, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.35473307967185974, + "learning_rate": 2.0145e-05, + "loss": 0.0503, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24238434433937073, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0511, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.32721346616744995, + "learning_rate": 2.0245e-05, + "loss": 0.0474, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.3787407875061035, + "learning_rate": 2.0295e-05, + "loss": 0.0515, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.3435424566268921, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0521, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.250515341758728, + "learning_rate": 2.0395e-05, + "loss": 0.0539, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.2730806767940521, + "learning_rate": 2.0445e-05, + "loss": 0.0544, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.32704228162765503, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0516, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.26865822076797485, + "learning_rate": 2.0545e-05, + "loss": 0.0557, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.31514087319374084, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.051, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.2616795301437378, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0483, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2979176938533783, + "learning_rate": 2.0695e-05, + "loss": 0.056, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.2876877188682556, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0496, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.2801620662212372, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0533, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.3521801233291626, + "learning_rate": 2.0845e-05, + "loss": 0.051, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.3016152083873749, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0515, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.29837727546691895, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0483, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.255982369184494, + "learning_rate": 2.0995e-05, + "loss": 0.0482, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.20852255821228027, + "learning_rate": 2.1045e-05, + "loss": 0.05, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.2746782898902893, + "learning_rate": 2.1095e-05, + "loss": 0.0513, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.28745001554489136, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0486, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.3163246810436249, + "learning_rate": 2.1195e-05, + "loss": 0.0536, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.2279432713985443, + "learning_rate": 2.1245e-05, + "loss": 0.0497, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.2273164689540863, + "learning_rate": 2.1295e-05, + "loss": 0.0485, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.3124007284641266, + "learning_rate": 2.1345e-05, + "loss": 0.0473, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.2097717970609665, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0514, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.24428582191467285, + "learning_rate": 2.1445e-05, + "loss": 0.0543, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.47236987948417664, + "learning_rate": 2.1495e-05, + "loss": 0.0525, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.3367164433002472, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0525, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.34086376428604126, + "learning_rate": 2.1595e-05, + "loss": 0.049, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.3060646951198578, + "learning_rate": 2.1645e-05, + "loss": 0.0538, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.2997030019760132, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0553, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.2652549743652344, + "learning_rate": 2.1745e-05, + "loss": 0.051, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.3072361350059509, + "learning_rate": 2.1795e-05, + "loss": 0.0516, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2494215965270996, + "learning_rate": 2.1845e-05, + "loss": 0.0476, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.20796196162700653, + "learning_rate": 2.1895e-05, + "loss": 0.0553, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.24235095083713531, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.21033242344856262, + "learning_rate": 2.1995e-05, + "loss": 0.0542, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.26056164503097534, + "learning_rate": 2.2045e-05, + "loss": 0.0546, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.24798662960529327, + "learning_rate": 2.2095e-05, + "loss": 0.0529, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.22179707884788513, + "learning_rate": 2.2145e-05, + "loss": 0.0542, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.2660025358200073, + "learning_rate": 2.2195e-05, + "loss": 0.0496, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.3771500289440155, + "learning_rate": 2.2245e-05, + "loss": 0.0502, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.3059975206851959, + "learning_rate": 2.2295e-05, + "loss": 0.0482, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.2941029667854309, + "learning_rate": 2.2345e-05, + "loss": 0.0483, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3481833040714264, + "learning_rate": 2.2395e-05, + "loss": 0.0541, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.2729704678058624, + "learning_rate": 2.2445e-05, + "loss": 0.0514, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.44202038645744324, + "learning_rate": 2.2495e-05, + "loss": 0.0534, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.311088889837265, + "learning_rate": 2.2545e-05, + "loss": 0.0474, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30663055181503296, + "learning_rate": 2.2595e-05, + "loss": 0.0491, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.3530942499637604, + "learning_rate": 2.2645e-05, + "loss": 0.0503, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.27998727560043335, + "learning_rate": 2.2695e-05, + "loss": 0.0493, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.2679221034049988, + "learning_rate": 2.2745e-05, + "loss": 0.0512, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.348821222782135, + "learning_rate": 2.2795e-05, + "loss": 0.0489, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.260958194732666, + "learning_rate": 2.2845e-05, + "loss": 0.0474, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.2429468184709549, + "learning_rate": 2.2895e-05, + "loss": 0.053, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.2635575234889984, + "learning_rate": 2.2945e-05, + "loss": 0.053, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.3416514992713928, + "learning_rate": 2.2995e-05, + "loss": 0.0476, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.221172496676445, + "learning_rate": 2.3045e-05, + "loss": 0.0519, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.30541878938674927, + "learning_rate": 2.3095e-05, + "loss": 0.0522, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.25274574756622314, + "learning_rate": 2.3145e-05, + "loss": 0.0499, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.366253525018692, + "learning_rate": 2.3195e-05, + "loss": 0.0463, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.16825024783611298, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0493, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.3158467710018158, + "learning_rate": 2.3295e-05, + "loss": 0.0482, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.2887287139892578, + "learning_rate": 2.3345e-05, + "loss": 0.053, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.4717058539390564, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0515, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.3148113489151001, + "learning_rate": 2.3445e-05, + "loss": 0.0523, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.23815515637397766, + "learning_rate": 2.3495e-05, + "loss": 0.0523, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23897255957126617, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0484, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.2680213749408722, + "learning_rate": 2.3595e-05, + "loss": 0.0485, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.29999756813049316, + "learning_rate": 2.3645e-05, + "loss": 0.0506, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.30607151985168457, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.05, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.26077136397361755, + "learning_rate": 2.3745e-05, + "loss": 0.0558, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2638765275478363, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.049, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.24569877982139587, + "learning_rate": 2.3845e-05, + "loss": 0.053, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.29622989892959595, + "learning_rate": 2.3895e-05, + "loss": 0.0517, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.2598831355571747, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.05, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.25716227293014526, + "learning_rate": 2.3995e-05, + "loss": 0.0453, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.37305381894111633, + "learning_rate": 2.4045e-05, + "loss": 0.0539, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.3151630461215973, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0516, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.296941339969635, + "learning_rate": 2.4145e-05, + "loss": 0.0509, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.31995704770088196, + "learning_rate": 2.4195e-05, + "loss": 0.0498, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2426208257675171, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0475, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.23927678167819977, + "learning_rate": 2.4295e-05, + "loss": 0.05, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.1887904703617096, + "learning_rate": 2.4345e-05, + "loss": 0.05, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.18480589985847473, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0496, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.24515867233276367, + "learning_rate": 2.4445e-05, + "loss": 0.0464, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.25505101680755615, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0481, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.24435125291347504, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0502, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.25899314880371094, + "learning_rate": 2.4595e-05, + "loss": 0.0535, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.30518922209739685, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0481, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.26098036766052246, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0477, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.24723340570926666, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.26713746786117554, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0506, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.21166373789310455, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0502, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2602159082889557, + "learning_rate": 2.4895e-05, + "loss": 0.0506, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27918240427970886, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.049, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.2131885290145874, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0477, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.28778040409088135, + "learning_rate": 2.5045e-05, + "loss": 0.0487, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.3056010603904724, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.049, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.2850024700164795, + "learning_rate": 2.5145e-05, + "loss": 0.0475, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.20490600168704987, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0497, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.32371339201927185, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0491, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.32466208934783936, + "learning_rate": 2.5295e-05, + "loss": 0.0485, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.2779853343963623, + "learning_rate": 2.5345e-05, + "loss": 0.0485, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.3257012367248535, + "learning_rate": 2.5395e-05, + "loss": 0.0486, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2625649571418762, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0486, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.19846542179584503, + "learning_rate": 2.5495e-05, + "loss": 0.0494, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.2415613830089569, + "learning_rate": 2.5545e-05, + "loss": 0.0539, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.20438073575496674, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0476, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.2543872892856598, + "learning_rate": 2.5645e-05, + "loss": 0.0489, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.26969635486602783, + "learning_rate": 2.5695e-05, + "loss": 0.0557, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.27818888425827026, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0517, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.28826993703842163, + "learning_rate": 2.5795e-05, + "loss": 0.0499, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.19518250226974487, + "learning_rate": 2.5845e-05, + "loss": 0.0499, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.2639068067073822, + "learning_rate": 2.5895e-05, + "loss": 0.0459, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.26014742255210876, + "learning_rate": 2.5945e-05, + "loss": 0.0522, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.26659756898880005, + "learning_rate": 2.5995e-05, + "loss": 0.0471, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.24073593318462372, + "learning_rate": 2.6045e-05, + "loss": 0.0488, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.2724616527557373, + "learning_rate": 2.6095e-05, + "loss": 0.0493, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.3012109398841858, + "learning_rate": 2.6145e-05, + "loss": 0.049, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23306235671043396, + "learning_rate": 2.6195e-05, + "loss": 0.0474, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.23879121243953705, + "learning_rate": 2.6245e-05, + "loss": 0.048, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20701922476291656, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0527, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.28734415769577026, + "learning_rate": 2.6345e-05, + "loss": 0.0469, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.26227426528930664, + "learning_rate": 2.6395e-05, + "loss": 0.0472, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.24196277558803558, + "learning_rate": 2.6445e-05, + "loss": 0.0462, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.3405964970588684, + "learning_rate": 2.6495e-05, + "loss": 0.0479, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.24121852219104767, + "learning_rate": 2.6545e-05, + "loss": 0.0444, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.284856379032135, + "learning_rate": 2.6595e-05, + "loss": 0.0492, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.2515311539173126, + "learning_rate": 2.6645e-05, + "loss": 0.0453, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.29297393560409546, + "learning_rate": 2.6695e-05, + "loss": 0.049, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.2528100311756134, + "learning_rate": 2.6745e-05, + "loss": 0.0477, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.29126009345054626, + "learning_rate": 2.6795e-05, + "loss": 0.0484, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.26489511132240295, + "learning_rate": 2.6845e-05, + "loss": 0.0476, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.33150023221969604, + "learning_rate": 2.6895e-05, + "loss": 0.0475, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.25904157757759094, + "learning_rate": 2.6945e-05, + "loss": 0.0495, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.24830900132656097, + "learning_rate": 2.6995e-05, + "loss": 0.0476, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.277997225522995, + "learning_rate": 2.7045e-05, + "loss": 0.0478, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.3100328743457794, + "learning_rate": 2.7095e-05, + "loss": 0.0501, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.2957077622413635, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0491, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.19402435421943665, + "learning_rate": 2.7195e-05, + "loss": 0.0426, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.20080482959747314, + "learning_rate": 2.7245e-05, + "loss": 0.0464, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.21148671209812164, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0457, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.2299162596464157, + "learning_rate": 2.7345e-05, + "loss": 0.0467, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2296467125415802, + "learning_rate": 2.7395e-05, + "loss": 0.0489, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.25099772214889526, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0487, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2503034472465515, + "learning_rate": 2.7495e-05, + "loss": 0.0533, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.24678738415241241, + "learning_rate": 2.7545e-05, + "loss": 0.0462, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.3281477987766266, + "learning_rate": 2.7595e-05, + "loss": 0.0492, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.32214412093162537, + "learning_rate": 2.7645e-05, + "loss": 0.0459, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.29244086146354675, + "learning_rate": 2.7695e-05, + "loss": 0.0495, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.2372455894947052, + "learning_rate": 2.7745e-05, + "loss": 0.0477, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.28411489725112915, + "learning_rate": 2.7795e-05, + "loss": 0.0514, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.36217001080513, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2913915514945984, + "learning_rate": 2.7895e-05, + "loss": 0.0535, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.3237503170967102, + "learning_rate": 2.7945e-05, + "loss": 0.0509, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.32016751170158386, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0526, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.25484171509742737, + "learning_rate": 2.8045e-05, + "loss": 0.0508, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.22641469538211823, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.23269280791282654, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0484, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23346202075481415, + "learning_rate": 2.8195e-05, + "loss": 0.0472, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.25522616505622864, + "learning_rate": 2.8245e-05, + "loss": 0.0493, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.2478165179491043, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0468, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.25732970237731934, + "learning_rate": 2.8345e-05, + "loss": 0.0484, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.21047812700271606, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0455, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.252676397562027, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0496, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20048962533473969, + "learning_rate": 2.8495e-05, + "loss": 0.0434, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.1821785420179367, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0502, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.24106337130069733, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0471, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.413974404335022, + "learning_rate": 2.8645e-05, + "loss": 0.0542, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.29419073462486267, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0499, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.20694036781787872, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0453, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2390667200088501, + "learning_rate": 2.8795e-05, + "loss": 0.045, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2563699781894684, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.046, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.22678422927856445, + "learning_rate": 2.8895e-05, + "loss": 0.0456, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.26104822754859924, + "learning_rate": 2.8945e-05, + "loss": 0.0446, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.22175569832324982, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0478, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.20990405976772308, + "learning_rate": 2.9045e-05, + "loss": 0.0459, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.27277135848999023, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0477, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.22422750294208527, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.047, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.2740926742553711, + "learning_rate": 2.9195e-05, + "loss": 0.0444, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2977927327156067, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0477, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26431694626808167, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0488, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.22759316861629486, + "learning_rate": 2.9345e-05, + "loss": 0.0465, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.24671563506126404, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0501, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.24065586924552917, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0439, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.3334764838218689, + "learning_rate": 2.9495e-05, + "loss": 0.0475, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.23565605282783508, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.046, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.26102808117866516, + "learning_rate": 2.9595e-05, + "loss": 0.0494, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.34544840455055237, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.05, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.2476256936788559, + "learning_rate": 2.9695e-05, + "loss": 0.0477, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.2795628309249878, + "learning_rate": 2.9745e-05, + "loss": 0.0508, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.2756715714931488, + "learning_rate": 2.9795e-05, + "loss": 0.0484, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.20578329265117645, + "learning_rate": 2.9845e-05, + "loss": 0.0461, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.20354008674621582, + "learning_rate": 2.9895e-05, + "loss": 0.0431, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24710936844348907, + "learning_rate": 2.9945e-05, + "loss": 0.0459, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.32639119029045105, + "learning_rate": 2.9995e-05, + "loss": 0.045, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.27762553095817566, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0476, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.2903423607349396, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.3292105793952942, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0503, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.23503656685352325, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0446, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.24153585731983185, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.048, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.24157844483852386, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.046, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.23378467559814453, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0478, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.20753100514411926, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0527, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.2574090361595154, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0468, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.23341813683509827, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0445, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21378666162490845, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0487, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.2410328984260559, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0488, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.22005778551101685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0478, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.3197212219238281, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.046, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.253745436668396, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0437, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.1991635262966156, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0474, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.20773734152317047, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0439, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.3026197552680969, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0462, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.22027666866779327, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0443, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.2512618899345398, + "learning_rate": 2.99997744429684e-05, + "loss": 0.044, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.23001381754875183, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0479, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.20057332515716553, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0433, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.2831740975379944, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0462, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.18362601101398468, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0472, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.29741403460502625, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0443, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.20039376616477966, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0457, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.14999084174633026, + "learning_rate": 2.999958785135414e-05, + "loss": 0.043, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.18134453892707825, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0416, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.25010257959365845, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0494, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.31700536608695984, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0455, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.25080761313438416, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0436, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.2008715718984604, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0459, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.21441133320331573, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0492, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.23232029378414154, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0459, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.30173608660697937, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0505, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.21179363131523132, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0438, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.21905283629894257, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0439, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.17932383716106415, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0452, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.21132917702198029, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0454, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.22732721269130707, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.047, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.28214913606643677, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0446, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.23979637026786804, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0454, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.3095552921295166, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0449, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.2585015594959259, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0473, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.18226519227027893, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0472, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2568792402744293, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0443, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.27914026379585266, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0423, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.21452564001083374, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0457, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.22575126588344574, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0458, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.24053922295570374, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0453, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.21956735849380493, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.048, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2325354665517807, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0439, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.1692194789648056, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.044, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.25796884298324585, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0416, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.2719244956970215, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0445, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.23312833905220032, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0488, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.24458816647529602, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0473, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.24380528926849365, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0411, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.1836114227771759, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0433, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.21051307022571564, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0462, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.18093563616275787, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.045, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2444952130317688, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0485, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.19429947435855865, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0487, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.2294744998216629, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0479, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.2147439569234848, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0492, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.2777695059776306, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0419, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22944948077201843, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0458, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.2822294235229492, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0429, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.20201396942138672, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0435, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.1961182802915573, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0441, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.1653209924697876, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0473, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.2082156240940094, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0508, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.199684277176857, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0466, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.20871016383171082, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0436, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.22781118750572205, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0504, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2652275562286377, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0467, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.20812813937664032, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0456, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2533996105194092, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0458, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.18139654397964478, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0441, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.27243369817733765, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0472, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.326709121465683, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0487, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.26698657870292664, + "learning_rate": 2.999617967180403e-05, + "loss": 0.05, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.4119804799556732, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0455, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.21175070106983185, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0462, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.22503416240215302, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.0468, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.254734605550766, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0471, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26713427901268005, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0439, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.2739362120628357, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0455, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.26221534609794617, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0451, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26752719283103943, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.23052464425563812, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0481, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.162479966878891, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0435, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.24716691672801971, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0459, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.14981329441070557, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0407, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.23326952755451202, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0459, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.2200203239917755, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.046, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.21501530706882477, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0433, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.22950918972492218, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0454, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.2042679488658905, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0444, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.16468611359596252, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0434, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.2180141657590866, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0447, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.1658746749162674, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.24921023845672607, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0398, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0445, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.25189560651779175, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0429, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.2614912688732147, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0427, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.2890806794166565, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0458, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.24215207993984222, + "learning_rate": 2.999336925698801e-05, + "loss": 0.04, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.1806204915046692, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0435, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.18884161114692688, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0422, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.2760990262031555, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0427, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.22660709917545319, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0459, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.24862384796142578, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.045, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.1911306381225586, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0433, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.2612020969390869, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0414, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.20945195853710175, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.043, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.2436099350452423, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0472, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.23788926005363464, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0455, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.20128341019153595, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0456, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.24002164602279663, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0414, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.2653037905693054, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.042, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.15122903883457184, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0395, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.19299687445163727, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0452, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.2534617781639099, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0445, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.29057401418685913, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0398, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.19744732975959778, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.041, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.25267109274864197, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0439, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2406993955373764, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0445, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.2561663091182709, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0465, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.2137032002210617, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0407, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.3002893030643463, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0433, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.32190272212028503, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0444, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.2076684534549713, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0432, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.31036099791526794, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0431, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.21386218070983887, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0442, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.19012001156806946, + "learning_rate": 2.998948184763285e-05, + "loss": 0.048, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.2310018688440323, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0466, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.1984279453754425, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0443, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.24101053178310394, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0446, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.17791084945201874, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0403, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.2016495168209076, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0404, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.19435155391693115, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.21197110414505005, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0398, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.22241441905498505, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.20825718343257904, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0434, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.24533888697624207, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.046, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.2455352544784546, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0412, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.18937505781650543, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0425, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.22356615960597992, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0418, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.23797748982906342, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0457, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.23130296170711517, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0427, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.20548853278160095, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0439, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19323743879795074, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0432, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.24339382350444794, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0412, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.18780124187469482, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0416, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.19353246688842773, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0461, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.21522872149944305, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0412, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.26847484707832336, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.3781932294368744, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0401, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.3219687342643738, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0423, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.1699237823486328, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0421, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.2174658179283142, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0398, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.2433294802904129, + "learning_rate": 2.998488804893431e-05, + "loss": 0.042, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.2116415947675705, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0405, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.22813065350055695, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0432, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.24252736568450928, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.048, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.2009168565273285, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0431, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.19379755854606628, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.23313777148723602, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0384, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.22056345641613007, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0459, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.3403800129890442, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0439, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.17649835348129272, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0419, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.20594412088394165, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0418, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.2901265323162079, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0405, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.22897225618362427, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0424, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.20961393415927887, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0438, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.28585514426231384, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0477, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2239968478679657, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0398, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.1651894897222519, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0405, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.1731943041086197, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0407, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.1708967089653015, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0417, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.17396977543830872, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0359, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.21327704191207886, + "learning_rate": 2.998095036004704e-05, + "loss": 0.045, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.15585461258888245, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0403, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.21774765849113464, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0403, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.1889292299747467, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0408, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.1639491617679596, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0425, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.23745876550674438, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.043, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19481611251831055, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0453, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.21001650393009186, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0429, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.2114192098379135, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0419, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.2506379783153534, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0392, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.21439281105995178, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0422, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.24763159453868866, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0433, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.1719428300857544, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0422, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.25943970680236816, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0425, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22226977348327637, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0448, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.2018195390701294, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0408, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.20869500935077667, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0443, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.23311494290828705, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.042, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.2501373589038849, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0382, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.2650911509990692, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0377, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.21386376023292542, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0405, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.27908873558044434, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0431, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.18731699883937836, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0402, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.2566622197628021, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0419, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.2418505847454071, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0406, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.199744313955307, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0424, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.20134931802749634, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.042, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.15607263147830963, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0453, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.22614715993404388, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0413, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.22810052335262299, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0433, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.16557516157627106, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0424, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.20254060626029968, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0394, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2068638950586319, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0408, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.27261367440223694, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0433, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.23216570913791656, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0445, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.2546656131744385, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0381, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.21829853951931, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0472, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.17109332978725433, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0396, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.1746513843536377, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0413, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.22411789000034332, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0373, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.21182285249233246, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.038, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.19430142641067505, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0419, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.22196920216083527, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0395, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.20695635676383972, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0407, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.1721169799566269, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0396, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.23676829040050507, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0396, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.3019765317440033, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0443, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.23958736658096313, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0402, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.17959944903850555, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0434, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.1563226878643036, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0397, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.16746941208839417, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0402, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.26673364639282227, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0411, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.23634585738182068, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0427, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.19724313914775848, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0408, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.2072090059518814, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0389, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.18574072420597076, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0427, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.23390066623687744, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0411, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.2137259542942047, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0395, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.21458280086517334, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0422, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.1838904619216919, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.043, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.1834891140460968, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0389, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.2012249082326889, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.27410435676574707, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0473, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.21607685089111328, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0405, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.17525449395179749, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.043, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.23137639462947845, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0414, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.20958152413368225, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0404, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.21453802287578583, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.043, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.2522047758102417, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.04, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.23595386743545532, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0396, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.23375281691551208, + "learning_rate": 2.996358559911323e-05, + "loss": 0.042, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.21628527343273163, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0434, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.2463727742433548, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0391, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.19919127225875854, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.043, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.28369203209877014, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0463, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.30222833156585693, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0397, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.19382458925247192, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0402, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2075910121202469, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0392, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.2503066658973694, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0427, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.18662676215171814, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0397, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.2432677149772644, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0425, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.19900213181972504, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0422, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.12742096185684204, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.20504114031791687, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0404, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.19127227365970612, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0425, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.18699967861175537, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0419, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.20074906945228577, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0381, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.24741096794605255, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0389, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.22948721051216125, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0435, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.17772004008293152, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0412, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.228739395737648, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0406, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.18312685191631317, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0401, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.21340692043304443, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0403, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.17110474407672882, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0388, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.19640640914440155, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.2386539727449417, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.043, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.20800922811031342, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0412, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1889747679233551, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0377, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.1801699548959732, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0402, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.20975525677204132, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0426, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.22626695036888123, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0404, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.2192756086587906, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0438, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.21307197213172913, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0447, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.28529244661331177, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0412, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.26367515325546265, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.044, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.29029133915901184, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.14019586145877838, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0427, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19176369905471802, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0408, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.19753597676753998, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0395, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1900634765625, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0427, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.21698379516601562, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0452, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.21722441911697388, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0417, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.1760135143995285, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0405, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18030588328838348, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0425, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.20307210087776184, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0416, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.19713939726352692, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0401, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.18719582259655, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.19560874998569489, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.04, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1633717119693756, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0424, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.20183435082435608, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0397, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.23784901201725006, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0438, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.17358708381652832, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0393, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18991878628730774, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0387, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.20455898344516754, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.042, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.23264537751674652, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0447, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2586922347545624, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0402, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.20447152853012085, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0406, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.1814388930797577, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0412, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2679784595966339, + "learning_rate": 2.994497917870469e-05, + "loss": 0.039, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.1818714141845703, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0423, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.2153957635164261, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0391, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.1858970820903778, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0394, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.23248426616191864, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.17587599158287048, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0367, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.18720309436321259, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.039, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.17602671682834625, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0393, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.15901489555835724, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0418, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.18207870423793793, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0401, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.22800953686237335, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0416, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.20031091570854187, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0417, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.24054989218711853, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0438, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.201461061835289, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0369, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.15697605907917023, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0397, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.16894802451133728, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0403, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2567543089389801, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0421, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.20297718048095703, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0396, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.17331846058368683, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.038, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.2214743047952652, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0405, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.2107590287923813, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0388, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.1768934726715088, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0364, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.16018562018871307, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0405, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.18402984738349915, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0425, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.21482129395008087, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0399, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.20554907619953156, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0388, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.2401154637336731, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0369, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.15072989463806152, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0359, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.1649719625711441, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0399, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.2386736422777176, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0403, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.2389005720615387, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0416, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.14930105209350586, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0372, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.18874907493591309, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0356, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.21466751396656036, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0391, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1697709560394287, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.24686385691165924, + "learning_rate": 2.993190369330068e-05, + "loss": 0.04, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.19733406603336334, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0417, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.3981468677520752, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0375, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.20926636457443237, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0384, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.18989330530166626, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0365, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17171184718608856, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0386, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.20474061369895935, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0413, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.19314105808734894, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0358, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.1794378012418747, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0409, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.17263059318065643, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0385, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.17863675951957703, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0355, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.20071984827518463, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0386, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.2025180608034134, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0409, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.21840184926986694, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0397, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.197452113032341, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0379, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.15370187163352966, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0379, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.22221392393112183, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0396, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.18536874651908875, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.18744619190692902, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0378, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.21197383105754852, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0399, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.29221004247665405, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0355, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.2411249279975891, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.039, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.3187369108200073, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.036, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.18096214532852173, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0372, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.2743191719055176, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0401, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.1864418387413025, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0384, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.15518276393413544, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.039, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.18089881539344788, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0359, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.2218618243932724, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0427, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.17051172256469727, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0401, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.15884852409362793, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0378, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1561235636472702, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0383, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.19711007177829742, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0409, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.18595683574676514, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.043, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.18613699078559875, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0357, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.1390007734298706, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0356, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.25603482127189636, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0377, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.1902502328157425, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0381, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.18781402707099915, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0393, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.20504912734031677, + "learning_rate": 2.991569760600352e-05, + "loss": 0.042, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.17359918355941772, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0373, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.2029414027929306, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0412, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.23485742509365082, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0398, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1454097479581833, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0394, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.22343166172504425, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0386, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.17754614353179932, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0358, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.14276635646820068, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0405, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.2062550187110901, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0429, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2093406766653061, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.038, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.14510402083396912, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0374, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.22726939618587494, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0373, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.14905233681201935, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0392, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.15828309953212738, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0361, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.990945968675917e-05, + "loss": 0.037, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.1867590844631195, + "learning_rate": 2.990900562792633e-05, + "loss": 0.038, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.18278446793556213, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0358, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.24628379940986633, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0393, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.2195759266614914, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0372, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.2397889941930771, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0352, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.1563243269920349, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0361, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.16278347373008728, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0366, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1776551604270935, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0369, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.19893717765808105, + "learning_rate": 2.990533239951709e-05, + "loss": 0.039, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.21403104066848755, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0378, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.19885864853858948, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0393, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.18425336480140686, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0404, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.1930677592754364, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0369, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18271560966968536, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0351, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.16518554091453552, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0367, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.17990152537822723, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0349, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.19625109434127808, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0368, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.1891077160835266, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0407, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.20162540674209595, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.17586444318294525, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0346, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.13146457076072693, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0338, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.16059468686580658, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0366, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.1724517047405243, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0395, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.1847251057624817, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0365, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.20104378461837769, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.037, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.2712373733520508, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0347, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.18014875054359436, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0388, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.17479945719242096, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0388, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.24176514148712158, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0387, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.2453809529542923, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0381, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2105131894350052, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0369, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1763688176870346, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0384, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.16223029792308807, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0348, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.173785999417305, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.038, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.23269419372081757, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0393, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.19176329672336578, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0399, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.1559731364250183, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0375, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.15701691806316376, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.22872254252433777, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0371, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15535064041614532, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0378, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.21613062918186188, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0382, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.1504523903131485, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0386, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.177171990275383, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2307550013065338, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0354, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.2215310037136078, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0371, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.23132716119289398, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0389, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.1937142014503479, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0373, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.23848143219947815, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0355, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.23107734322547913, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0359, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1455046832561493, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0404, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.21890339255332947, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0421, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.1897345334291458, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0357, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.2016463577747345, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0363, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.15850050747394562, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0391, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.18169008195400238, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0377, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.26095741987228394, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0391, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.20983633399009705, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0355, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.21026623249053955, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.2104070633649826, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0343, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1801508367061615, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0369, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.22245049476623535, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.0358, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.19220919907093048, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0354, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.2285826951265335, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0391, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.15382888913154602, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0374, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.13551367819309235, + "learning_rate": 2.987759237035982e-05, + "loss": 0.039, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.23362696170806885, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0367, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.15972478687763214, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0319, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.23858226835727692, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0354, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.17988596856594086, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0391, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.21262454986572266, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0392, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.1443355828523636, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0384, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.24133095145225525, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0382, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.175334632396698, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.039, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.20340824127197266, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0376, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.16549751162528992, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0371, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.16527467966079712, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0373, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1418737769126892, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0378, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18754976987838745, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0355, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.13649867475032806, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0329, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.21408453583717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0348, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.20929333567619324, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.24083013832569122, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0378, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17924576997756958, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0369, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.22360439598560333, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0382, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.15963956713676453, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0361, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.19003787636756897, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0365, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2458433210849762, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0375, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.22048939764499664, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0382, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.19965334236621857, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0371, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.228081613779068, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0375, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.24077889323234558, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0347, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.17692111432552338, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0365, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.1996857225894928, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0358, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.24799935519695282, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0377, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.24300269782543182, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0362, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.2230309098958969, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0364, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.16640354692935944, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0346, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.14670157432556152, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.039, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18531887233257294, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0353, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.19566190242767334, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0391, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.1967019885778427, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0355, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.2161232978105545, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0375, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.17630350589752197, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0397, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18905608355998993, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0347, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.1890578716993332, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.036, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.15397298336029053, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0401, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0362, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.20983757078647614, + "learning_rate": 2.985388676461079e-05, + "loss": 0.036, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.1554982215166092, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0378, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.1961488276720047, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0356, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.17265456914901733, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0357, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.1275556981563568, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0353, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.1667783409357071, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0354, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.27920690178871155, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0385, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.12451875954866409, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0352, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.15794622898101807, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0339, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.16285695135593414, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0369, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.18990163505077362, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0376, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.1867877095937729, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0363, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.15270273387432098, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0394, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.1934751272201538, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0382, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16943520307540894, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0369, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.20546840131282806, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0336, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.18490619957447052, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0415, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.16439616680145264, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0384, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.13623777031898499, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0379, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.18209043145179749, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0357, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.19922122359275818, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0365, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.15637168288230896, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0323, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1781233251094818, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0329, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.17547845840454102, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0344, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.1770639270544052, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0358, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.2677673101425171, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0356, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.16740617156028748, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0344, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.13076575100421906, + "learning_rate": 2.983793562696423e-05, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.17586535215377808, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0381, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.1623464673757553, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.037, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.16550037264823914, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0373, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.20314498245716095, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0366, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.19387134909629822, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0377, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.17531387507915497, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0381, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.21466659009456635, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.037, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1693795919418335, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0339, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.21929262578487396, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0332, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.2253660261631012, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0339, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.2084304839372635, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0363, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.2513981759548187, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0357, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.19585898518562317, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0353, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.22262930870056152, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0357, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.19813011586666107, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0347, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.16626323759555817, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0346, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.271424800157547, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.036, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2283792346715927, + "learning_rate": 2.982684510938631e-05, + "loss": 0.04, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.15996390581130981, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0394, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.20269536972045898, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0387, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.20575514435768127, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0386, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.15178649127483368, + "learning_rate": 2.982433099619666e-05, + "loss": 0.037, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.23187878727912903, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0355, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.12988153100013733, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0369, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19367633759975433, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0367, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.15319304168224335, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0337, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.19311629235744476, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0383, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.14771778881549835, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0393, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.20381954312324524, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0414, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22292324900627136, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0358, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.13758333027362823, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0359, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.18837112188339233, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0339, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.1757460981607437, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0352, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.24256370961666107, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0346, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.2417164295911789, + "learning_rate": 2.981603574229741e-05, + "loss": 0.037, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.21129392087459564, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0371, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16878797113895416, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0367, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.17526096105575562, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0353, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.1812993884086609, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.037, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.16877815127372742, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0353, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.20823006331920624, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0384, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1536571979522705, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0356, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.1477498710155487, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0346, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.12916848063468933, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0318, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.15418170392513275, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0369, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.1883322149515152, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0369, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.1569332480430603, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0374, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.21140694618225098, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0365, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.26859667897224426, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0395, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.2182101309299469, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0382, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.18896763026714325, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0372, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.17398814857006073, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0338, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.14043423533439636, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0361, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.14995983242988586, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0363, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.15542328357696533, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0361, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.16333894431591034, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0337, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.18489807844161987, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0352, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.1794673502445221, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.037, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.1882123351097107, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0414, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.16483181715011597, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.04, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.1727719008922577, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0361, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.1968923956155777, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0389, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1838955134153366, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.034, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.23556186258792877, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0369, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.2518751323223114, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0331, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18682284653186798, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0371, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.2049458622932434, + "learning_rate": 2.979480675356209e-05, + "loss": 0.04, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.11846950650215149, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0347, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.21428292989730835, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0354, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.154558002948761, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0359, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1178995817899704, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0364, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.16926616430282593, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0373, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.1721247136592865, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0346, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.1975203901529312, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0351, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.17340779304504395, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0347, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.22424420714378357, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0384, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.1688208281993866, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.16542011499404907, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0343, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.15097306668758392, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0372, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.15530438721179962, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0345, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.16840992867946625, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.034, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.19959625601768494, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0375, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.15717248618602753, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0344, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.21473488211631775, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.04, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.18634898960590363, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0346, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.16032545268535614, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0377, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.19153323769569397, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0408, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.16018801927566528, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0334, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.1498652696609497, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0349, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.1652575582265854, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0344, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.15636402368545532, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0331, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.22963644564151764, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0359, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.16931135952472687, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0389, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.17646029591560364, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0322, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.19198739528656006, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0318, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.16826699674129486, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0389, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.1693607121706009, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.037, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.147409588098526, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0343, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.1984914094209671, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0329, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.18371893465518951, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0357, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.16279958188533783, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0377, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2122160941362381, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0367, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.20099543035030365, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0343, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.17249372601509094, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0325, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1106863021850586, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0336, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.2338913530111313, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0331, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.1545400768518448, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0321, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.17114387452602386, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0323, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.16218030452728271, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.037, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.14618389308452606, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0306, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.17804935574531555, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0348, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1512002944946289, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0318, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.14877542853355408, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.037, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.1693819761276245, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0325, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.17713536322116852, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0356, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.1719064563512802, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0358, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.19448526203632355, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0363, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.18047156929969788, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0326, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.20896470546722412, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0362, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.15999653935432434, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0318, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.17288383841514587, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0336, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.21483922004699707, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0367, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.19133612513542175, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0336, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.16010481119155884, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0377, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.1778886914253235, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0354, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.14997199177742004, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0326, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.16331784427165985, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0326, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.18247577548027039, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0367, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1838751882314682, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0338, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.17194764316082, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0365, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.1354181170463562, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0315, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.19473528861999512, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0345, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14786818623542786, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0354, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.1931883543729782, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0345, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.159915491938591, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0357, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14373569190502167, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0309, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.1468040943145752, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0339, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.1592605859041214, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0374, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.17073556780815125, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0365, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.13732466101646423, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0331, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.18718381226062775, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0337, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.179783895611763, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0383, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.15795794129371643, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0368, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.13824236392974854, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0343, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.2081076204776764, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0385, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.1407512128353119, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0338, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.1619846224784851, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0325, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.184702068567276, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0353, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.18360818922519684, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0349, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.17683662474155426, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0332, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.17690211534500122, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0364, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.18250274658203125, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0336, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.12902002036571503, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0344, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.1598006635904312, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0332, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.17691342532634735, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.037, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.2213180959224701, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0324, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.14762577414512634, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0321, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.14543993771076202, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.034, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.18970227241516113, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0377, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.15049827098846436, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0341, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.1485714316368103, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0361, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1893187314271927, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0354, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.2345002293586731, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0336, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.24119973182678223, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0325, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.18577414751052856, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0337, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.1941320151090622, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0355, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.17646151781082153, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0342, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.13801687955856323, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0351, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.15124225616455078, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0337, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19114945828914642, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.033, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.23505449295043945, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0357, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.15116260945796967, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.15656375885009766, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0322, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.18982963263988495, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0368, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.18300476670265198, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0379, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.16485247015953064, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0335, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.19399121403694153, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0353, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.22944431006908417, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0343, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.16755999624729156, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0351, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.15779320895671844, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0318, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.2113935351371765, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0343, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.14698673784732819, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0332, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19958600401878357, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0357, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20250289142131805, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0334, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.1955910474061966, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0381, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.13143742084503174, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0321, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.15840105712413788, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0333, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.20588421821594238, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0347, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.1881185919046402, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0345, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.16714425384998322, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.032, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.16526015102863312, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0316, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.1865321844816208, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0351, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.19202300906181335, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0325, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.15551044046878815, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0339, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.15684190392494202, + "learning_rate": 2.969840338753429e-05, + "loss": 0.034, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.16806578636169434, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0319, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.1554284542798996, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0348, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1836974322795868, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0375, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.20003411173820496, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0364, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.1747797578573227, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0328, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.2204974740743637, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0331, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.17183512449264526, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0381, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.14398148655891418, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.03, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.17428989708423615, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0329, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.13864420354366302, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0319, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.15052364766597748, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0357, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.21513856947422028, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0349, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19239531457424164, + "learning_rate": 2.968758727655112e-05, + "loss": 0.0377, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.22284342348575592, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0358, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.17206035554409027, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0349, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.1991468220949173, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0316, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.13542494177818298, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0329, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.14473386108875275, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0346, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.1972355991601944, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0348, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.17447993159294128, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0337, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.14419333636760712, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0355, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.19937965273857117, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0381, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.18988190591335297, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0331, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.1552465558052063, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0333, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.1769426017999649, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0344, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.17652767896652222, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0354, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1598699688911438, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0354, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.17556554079055786, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0372, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.18424467742443085, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0372, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.1591036170721054, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.038, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.1973133534193039, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0332, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.139867901802063, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0336, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.17536301910877228, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0339, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16841425001621246, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0319, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.24087557196617126, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.1419876366853714, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0341, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0315, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.14541317522525787, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0326, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.18411535024642944, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0382, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1769087016582489, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0355, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.16459472477436066, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0354, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.1770380288362503, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.034, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.16219645738601685, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0316, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.18393918871879578, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0366, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.1436525285243988, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0318, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.18282611668109894, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0376, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.16712592542171478, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0367, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.13408221304416656, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0305, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.17664852738380432, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0312, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.13730520009994507, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0323, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1278112679719925, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0322, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.1742715984582901, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0339, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.15022823214530945, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0348, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.17816156148910522, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0348, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.18016202747821808, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0357, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.21158602833747864, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0331, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1747511923313141, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0301, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.16840165853500366, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0363, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.1507093459367752, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0352, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.2081453651189804, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.035, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.1459163874387741, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0331, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.17468400299549103, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0339, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.15824441611766815, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0309, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.19502659142017365, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0305, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.1502636969089508, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.032, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.13870884478092194, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0356, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1596919149160385, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0347, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.1830880492925644, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0359, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.13567551970481873, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0319, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.12907959520816803, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0327, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.17357082664966583, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.0308, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.20002517104148865, + "learning_rate": 2.963613172841619e-05, + "loss": 0.035, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.13852699100971222, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0353, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.15445692837238312, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.11687827110290527, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0319, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.15454575419425964, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0366, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.1916663944721222, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0327, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.16530223190784454, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0309, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17331530153751373, + "learning_rate": 2.962976980219581e-05, + "loss": 0.032, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.17225238680839539, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0323, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.16390499472618103, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0324, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.16459178924560547, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0315, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.1416798084974289, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0348, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.1546057015657425, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0329, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15206973254680634, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0304, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.18791860342025757, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.03, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.17669418454170227, + "learning_rate": 2.962243236871436e-05, + "loss": 0.034, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.1335674524307251, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0323, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.17903365194797516, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0327, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.17771181464195251, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0341, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1532796174287796, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0328, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.18777944147586823, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0343, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.2553282082080841, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0342, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.24115385115146637, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0347, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.18773671984672546, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0324, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.18168678879737854, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0355, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.14390279352664948, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0323, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.13260799646377563, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0336, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.16469483077526093, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0341, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.1618199497461319, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0305, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.15524879097938538, + "learning_rate": 2.960942086834339e-05, + "loss": 0.034, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.14689622819423676, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0352, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.17949312925338745, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0329, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.2376648485660553, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0326, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.1302678883075714, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0356, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.16316285729408264, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0318, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.14314667880535126, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0347, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.13162434101104736, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0335, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.13411720097064972, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0334, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.20853494107723236, + "learning_rate": 2.960094148810064e-05, + "loss": 0.036, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.15634609758853912, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0314, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.16109681129455566, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0323, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.15969422459602356, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0292, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.13126519322395325, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0338, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.13577662408351898, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0323, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13226218521595, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0307, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.12283602356910706, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0305, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16819486021995544, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0316, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.1443309634923935, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.031, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.17339679598808289, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0335, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.12366446852684021, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0319, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.18073806166648865, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0355, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.16230227053165436, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.033, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.15816937386989594, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0331, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.15911494195461273, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0296, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.19500406086444855, + "learning_rate": 2.958564529914256e-05, + "loss": 0.034, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.16553588211536407, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0348, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.1780945360660553, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0297, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16839291155338287, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0317, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.26763999462127686, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0335, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.1292608082294464, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0319, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19598962366580963, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0346, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.16073159873485565, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0317, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.2037622481584549, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0324, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.11946277320384979, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0329, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.12799547612667084, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0329, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.23773977160453796, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0327, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.19241386651992798, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0371, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.15968935191631317, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0335, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.1712830811738968, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0302, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.16464142501354218, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0346, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.2122460901737213, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0319, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.16571341454982758, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.033, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1610543429851532, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0347, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14235250651836395, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0322, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.14999081194400787, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0299, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.033, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.1448807269334793, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0307, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.20123368501663208, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0315, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.16191548109054565, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.031, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.1475522518157959, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.17617955803871155, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0323, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.20633190870285034, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0287, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.13562145829200745, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0347, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.15800844132900238, + "learning_rate": 2.955719840324779e-05, + "loss": 0.034, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.1652759164571762, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0291, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.11819476634263992, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0309, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.17562547326087952, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0327, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.1934351772069931, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0342, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.1565367877483368, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.033, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.21048930287361145, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0357, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.16762706637382507, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0341, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.1267809122800827, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0322, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.18461033701896667, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0357, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.13451576232910156, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0327, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.15478897094726562, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0311, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1461619883775711, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0314, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.1432616114616394, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0339, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.14061163365840912, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0337, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.1950581669807434, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0315, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.13942912220954895, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.031, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.15901322662830353, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0324, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.19862298667430878, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0322, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.15155215561389923, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0302, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16394968330860138, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0299, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.13225089013576508, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0305, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.18203529715538025, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0319, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.15323573350906372, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0346, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.18822363018989563, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.038, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1510896533727646, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0336, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.12274178862571716, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.032, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.1287696361541748, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0341, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.16619601845741272, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0321, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.12426833063364029, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0298, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.1751766800880432, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0341, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.13578996062278748, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0306, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.19475996494293213, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0314, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.13291390240192413, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0341, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.16193236410617828, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0325, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.16108311712741852, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0356, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.16233862936496735, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0303, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.14042457938194275, + "learning_rate": 2.951955450025455e-05, + "loss": 0.031, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.14922280609607697, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0336, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.1389254927635193, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0324, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.21042059361934662, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0331, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.18529878556728363, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0311, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1345708668231964, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0316, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.14794935286045074, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0367, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.18609392642974854, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0326, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.1762595772743225, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0312, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.15549533069133759, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0342, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.1910371482372284, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0316, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.12789981067180634, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0286, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.11457879841327667, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0303, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.12534750998020172, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0313, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.17089471220970154, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0318, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.16774669289588928, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0326, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.18693485856056213, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0356, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.20419354736804962, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15851624310016632, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0328, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1284184604883194, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.1148369237780571, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0291, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.15302640199661255, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.037, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.16082480549812317, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.031, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.16006684303283691, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0346, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14866025745868683, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0323, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.15490789711475372, + "learning_rate": 2.949326462724779e-05, + "loss": 0.033, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.18960629403591156, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0302, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.16329076886177063, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0338, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1700190156698227, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0326, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.14541998505592346, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0332, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.13809162378311157, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.16794940829277039, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0345, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1897762268781662, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0304, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2064688503742218, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0332, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.14193041622638702, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0332, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.15571711957454681, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0339, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.14264710247516632, + "learning_rate": 2.948147907679037e-05, + "loss": 0.036, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.1411963850259781, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0344, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.18891720473766327, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0308, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.15993206202983856, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0339, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.15609434247016907, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0315, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.19272638857364655, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0325, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.2836095690727234, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0323, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16382543742656708, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0335, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.18899215757846832, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0301, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.17764721810817719, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0322, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.1808568239212036, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0304, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.166172593832016, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0298, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.18262767791748047, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0305, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.17854458093643188, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.033, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.16440987586975098, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.033, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.19478020071983337, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0288, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.2605016231536865, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0318, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1850014477968216, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0313, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.1339086890220642, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.17828164994716644, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0315, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.1235261932015419, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0328, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.13704872131347656, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0322, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1296597719192505, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0315, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.16906750202178955, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0347, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14643864333629608, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.031, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.16102853417396545, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0315, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.17575663328170776, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0312, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1488381028175354, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0318, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.14629733562469482, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0333, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.17051763832569122, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.034, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.14636048674583435, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0315, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.19957765936851501, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.032, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.15027561783790588, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0342, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1580146700143814, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.032, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1452188491821289, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0302, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.1924576759338379, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0323, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.199595108628273, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0326, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.1557701975107193, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0303, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.21232250332832336, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0324, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.13840343058109283, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0321, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.14219941198825836, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0314, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1747855842113495, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0342, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1730964183807373, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0334, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.12938249111175537, + "learning_rate": 2.94341324778745e-05, + "loss": 0.029, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1782408058643341, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0299, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.943188093053675e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.12131186574697495, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0277, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.19141671061515808, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0307, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.2431664764881134, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0326, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15145368874073029, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0329, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.19781085848808289, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0332, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.19477561116218567, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0317, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.1818031370639801, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0352, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.13725785911083221, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0272, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.1874770075082779, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0322, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.18701452016830444, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0304, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.15173715353012085, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.032, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.1315767467021942, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0331, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.16627946496009827, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0299, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.2122701108455658, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0308, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.12093837559223175, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0303, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.2072226107120514, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0331, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.14395366609096527, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0323, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.18454021215438843, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0349, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1647966057062149, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0342, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1907777190208435, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0347, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.14438189566135406, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0295, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.16802345216274261, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0283, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.1488419622182846, + "learning_rate": 2.940567320758981e-05, + "loss": 0.033, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14881044626235962, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0318, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.18762606382369995, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0322, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.15426765382289886, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0314, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.17882433533668518, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0361, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.16833186149597168, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0319, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.16899487376213074, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0283, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.15301142632961273, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0325, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.13270439207553864, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0305, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.16603541374206543, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0347, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.21959784626960754, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0299, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.24568438529968262, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0339, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.14889934659004211, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0284, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.20354624092578888, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0309, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.18124797940254211, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0352, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.13871392607688904, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0311, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.18997471034526825, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0299, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.09903991222381592, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0304, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.13336266577243805, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0306, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.1564595103263855, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0295, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.17495550215244293, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.17048844695091248, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0266, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.1257491409778595, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0299, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.15244479477405548, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0341, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.15737923979759216, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.03, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.13226579129695892, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0268, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.21549297869205475, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0323, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.2102872133255005, + "learning_rate": 2.937416924082395e-05, + "loss": 0.029, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.14369380474090576, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0306, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.1390914022922516, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0273, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.13536471128463745, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0299, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.13321688771247864, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0298, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13358351588249207, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0309, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.1732599437236786, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0319, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.17455634474754333, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0314, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.1714484691619873, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0345, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.18120428919792175, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0293, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.21681863069534302, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0335, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.12099924683570862, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0299, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.13233523070812225, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0372, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.16708864271640778, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0305, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.2122814804315567, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0319, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14159013330936432, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0343, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.13396364450454712, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0308, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.17630934715270996, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0349, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.16417403519153595, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0316, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.16593796014785767, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0291, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.13613900542259216, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0299, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.15980054438114166, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0319, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1578991562128067, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0288, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.1628173291683197, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0317, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.15937024354934692, + "learning_rate": 2.934549758921277e-05, + "loss": 0.031, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.15258124470710754, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0323, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.16142596304416656, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0289, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.15643659234046936, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0295, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.17128989100456238, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0313, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.1871197372674942, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0314, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.1448175013065338, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0318, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.21338209509849548, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0334, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15262271463871002, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0293, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.14302000403404236, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0307, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.13756762444972992, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0288, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.1390128880739212, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0286, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.16088609397411346, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0299, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.11852050572633743, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0284, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.17093200981616974, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0331, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.2325763702392578, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0269, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.19948336482048035, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0314, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.16918544471263885, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0311, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.14563684165477753, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0301, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.19645430147647858, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0312, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.18950732052326202, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0292, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.11659497767686844, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0307, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1314595341682434, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0326, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1457567662000656, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.16525784134864807, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0322, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.15977199375629425, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0334, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.17081087827682495, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0308, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.20844970643520355, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.029, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.22166086733341217, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1766587346792221, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0298, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1307239532470703, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0298, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0289, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.17264464497566223, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0317, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.15076857805252075, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0308, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14443998038768768, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0291, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.13115015625953674, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0329, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1444655954837799, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0294, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.16740760207176208, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0318, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.14201150834560394, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0314, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.14225733280181885, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0297, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.16758595407009125, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0307, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.16751281917095184, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.03, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.1855877935886383, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0333, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.17420977354049683, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0277, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.1568649560213089, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0312, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.15874658524990082, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0297, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.13727585971355438, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0334, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.14999760687351227, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0328, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.16789689660072327, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.14890174567699432, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0312, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.18001733720302582, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0322, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.1718987226486206, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.03, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.18727312982082367, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0297, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.1528918594121933, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0296, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.13372628390789032, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0288, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.14980314671993256, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0319, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.11368083208799362, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0312, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.14968374371528625, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0311, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.15593667328357697, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0287, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.1762111783027649, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0296, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.15493011474609375, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0272, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.14142747223377228, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.03, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.1307222545146942, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0294, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.1598598062992096, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.03, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.17966923117637634, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0337, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15533174574375153, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0276, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.1223452091217041, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0294, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.12733329832553864, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0289, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.15903618931770325, + "learning_rate": 2.926085774098786e-05, + "loss": 0.029, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1539735347032547, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0272, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.20487286150455475, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0312, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.19920656085014343, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0315, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.15179695188999176, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0305, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.16460831463336945, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0315, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.21031928062438965, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0283, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.13835182785987854, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0287, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.1597253829240799, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0336, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.11488606780767441, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0328, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.18707934021949768, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0328, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.18874959647655487, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0317, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.1182878389954567, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0316, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.11695742607116699, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0308, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.1597018986940384, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0324, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.17729975283145905, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0316, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.18250048160552979, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0309, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.20568659901618958, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0293, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.12907297909259796, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0281, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.1809365600347519, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0333, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.17379151284694672, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0327, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.16082972288131714, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0308, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.17415443062782288, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0299, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.1659587323665619, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0272, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.1538572907447815, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0309, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.16303570568561554, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0322, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1444631963968277, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0306, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.24458004534244537, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0285, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.15384158492088318, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0277, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1425933837890625, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0303, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.12485234439373016, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0279, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.14287105202674866, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0297, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.1621903032064438, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0268, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.145540252327919, + "learning_rate": 2.921797581830774e-05, + "loss": 0.03, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.1410767287015915, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0309, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.10953503102064133, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0299, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.15029482543468475, + "learning_rate": 2.92140191009304e-05, + "loss": 0.031, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.13262124359607697, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0299, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1503380835056305, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0329, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.18932904303073883, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.028, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.1473148614168167, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0315, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.18515919148921967, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0309, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.16122235357761383, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0279, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.1684419959783554, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0309, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.21400189399719238, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0304, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.12552157044410706, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0286, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.15383073687553406, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.029, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.1269458532333374, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0289, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.1839684545993805, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0316, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.13441112637519836, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0281, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.13045959174633026, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0296, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.13933542370796204, + "learning_rate": 2.919408984114327e-05, + "loss": 0.029, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.15143601596355438, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0296, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.16731251776218414, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.032, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.13245129585266113, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0268, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.17244738340377808, + "learning_rate": 2.918873439576597e-05, + "loss": 0.028, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.1709858626127243, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0293, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18727441132068634, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0302, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1482084095478058, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0297, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.1343366652727127, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0303, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.1456393450498581, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0299, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1569104641675949, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0281, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.19711613655090332, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0298, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.14027468860149384, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.031, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.16852174699306488, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0289, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.18446287512779236, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0308, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.16207733750343323, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0283, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.19747264683246613, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0301, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.15206509828567505, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0295, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.18873688578605652, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0286, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.144051656126976, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0335, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.14620175957679749, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0284, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.19520114362239838, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0304, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.16127106547355652, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0333, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.11467311531305313, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0303, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11777719110250473, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0311, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.12151405215263367, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0278, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1852089911699295, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0297, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.19555963575839996, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0321, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.1584533303976059, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0292, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.1764054298400879, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0306, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.16257385909557343, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0298, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.13896049559116364, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0296, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.16509538888931274, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0321, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.14912301301956177, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0285, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.18058481812477112, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0302, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.18734610080718994, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0289, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.19707795977592468, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0327, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.16215947270393372, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0302, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1710345447063446, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0332, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.19638675451278687, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0316, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.17807847261428833, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0285, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.13119569420814514, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0298, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.13414712250232697, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0301, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.1941734254360199, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0327, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.19624412059783936, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0291, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.17615875601768494, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0293, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.19011878967285156, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0316, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.15080299973487854, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0263, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.1676572859287262, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0318, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.14975300431251526, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0263, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17691564559936523, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.17775340378284454, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0298, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.14438258111476898, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.14729741215705872, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0289, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.20924119651317596, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0297, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.15471352636814117, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0289, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.1571032702922821, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0307, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15048687160015106, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0301, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.15063059329986572, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0279, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15528838336467743, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0299, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.1756942719221115, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0291, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.1491880565881729, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0296, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.1562451720237732, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0302, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.12366753816604614, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0257, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.15308785438537598, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.16614912450313568, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0301, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.12428635358810425, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0303, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.135500967502594, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0283, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.16944879293441772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0279, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.174363911151886, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0295, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.16406473517417908, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0284, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.2144644856452942, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0277, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.22006124258041382, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0264, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.15261462330818176, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0298, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.15205760300159454, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.14101643860340118, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0305, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16015839576721191, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.17772945761680603, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0312, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16818812489509583, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0283, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1448637843132019, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0325, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.14702047407627106, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0303, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.15547925233840942, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0299, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.18527188897132874, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.029, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.1533391922712326, + "learning_rate": 2.907944011202334e-05, + "loss": 0.029, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.15355326235294342, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0286, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.1503538340330124, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0312, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.12572433054447174, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0277, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.13614071905612946, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0303, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.13363786041736603, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0292, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.13318102061748505, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.0289, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.14687903225421906, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0301, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.13144652545452118, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0271, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.16351011395454407, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0292, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.19256997108459473, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0268, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13006120920181274, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0285, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.12932874262332916, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.11981740593910217, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.14302268624305725, + "learning_rate": 2.905937392537971e-05, + "loss": 0.029, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.1446913331747055, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0289, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.1500432789325714, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.15604284405708313, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0283, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.1795264482498169, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0289, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.14356553554534912, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0304, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.1323058009147644, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0302, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.2098175585269928, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0297, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.1424631029367447, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0295, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.1956447958946228, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0294, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.11993541568517685, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0273, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.18414588272571564, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0316, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.17320625483989716, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0319, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.17650462687015533, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0304, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.18740026652812958, + "learning_rate": 2.903909846652783e-05, + "loss": 0.034, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.12898528575897217, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0309, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.11721221357584, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0297, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.1659565567970276, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0294, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.1469241976737976, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0298, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14588896930217743, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0332, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.16243143379688263, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0305, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.15857404470443726, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0318, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.14992372691631317, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0305, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.13760966062545776, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0299, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.14302772283554077, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0313, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.18560320138931274, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0262, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.13117150962352753, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0286, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.16370882093906403, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0298, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.18689852952957153, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.031, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.16412676870822906, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0289, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.20153313875198364, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0297, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.15255704522132874, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0316, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.2036312222480774, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0306, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.16087041795253754, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0315, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.1514595001935959, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0295, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.13961558043956757, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0259, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1996241807937622, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0336, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.15389376878738403, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0278, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.12345903366804123, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.032, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.16360828280448914, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0292, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.19773486256599426, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.032, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.11931311339139938, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0268, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.1378420740365982, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0296, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.09076640754938126, + "learning_rate": 2.899643488777395e-05, + "loss": 0.026, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.1116260215640068, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0266, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.18670536577701569, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0314, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.15351636707782745, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0279, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.15310993790626526, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0298, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.15458840131759644, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0329, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.12034284323453903, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0266, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.14272604882717133, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0296, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.16335424780845642, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0266, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.16105882823467255, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0355, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.1455371230840683, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0261, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.21555180847644806, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0301, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.15278705954551697, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0292, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.14099596440792084, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0287, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.10848120599985123, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.028, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.1380046308040619, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0273, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.14006251096725464, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0328, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.16518674790859222, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0286, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.13295377790927887, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.031, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.1623251736164093, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0281, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.16029980778694153, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.03, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.15575948357582092, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0294, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.157670795917511, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0294, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1470523178577423, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0281, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.12192463129758835, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0248, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.1240374892950058, + "learning_rate": 2.895893835994484e-05, + "loss": 0.028, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1385754942893982, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0287, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.140727236866951, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0281, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.15758906304836273, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0294, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.19664499163627625, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0277, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.12567110359668732, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0279, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.13595536351203918, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0313, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.14262230694293976, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0317, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1474338322877884, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0282, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.18601401150226593, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0274, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.1392786204814911, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0269, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.12022347748279572, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0262, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.17340584099292755, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0277, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.1438349187374115, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0273, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.1408676654100418, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0277, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.14853249490261078, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0291, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.20401965081691742, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.031, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.18385861814022064, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.12695786356925964, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0267, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.14115597307682037, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0291, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.15790492296218872, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0294, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.15104074776172638, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0278, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13686735928058624, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0276, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.12029985338449478, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0305, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1351509988307953, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0293, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.10609307885169983, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0285, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15056312084197998, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0292, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.17062142491340637, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0281, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.15565407276153564, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0298, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.136429563164711, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0282, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.1649797409772873, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0267, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.19815988838672638, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.032, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15660062432289124, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.03, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.1840171217918396, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0342, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.14063017070293427, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0282, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.14935986697673798, + "learning_rate": 2.890533053632388e-05, + "loss": 0.029, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.14691074192523956, + "learning_rate": 2.89037798520224e-05, + "loss": 0.033, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.14992092549800873, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0262, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.16885656118392944, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0301, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1223178282380104, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0275, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.1200852021574974, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0277, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.1251501888036728, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0252, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.16905754804611206, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0277, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.13312722742557526, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0296, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.13198228180408478, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0269, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.14326612651348114, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0253, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.14861920475959778, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0262, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.19368910789489746, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0295, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.143910750746727, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0282, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.12314137071371078, + "learning_rate": 2.888352491214935e-05, + "loss": 0.027, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.16095006465911865, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0272, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.1414606273174286, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.029, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.15631091594696045, + "learning_rate": 2.887882538107974e-05, + "loss": 0.0291, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16476406157016754, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0283, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.2236901968717575, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0286, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.1318763792514801, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0274, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.15661245584487915, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.029, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.13406990468502045, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0287, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16861270368099213, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0272, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.18082283437252045, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0285, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.17517460882663727, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0277, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.17011325061321259, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0325, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14378181099891663, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0306, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1766553819179535, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0304, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.1666528880596161, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0294, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.15178363025188446, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0273, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.14275510609149933, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0291, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.1568918377161026, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0267, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.1370164304971695, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0268, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.10388749092817307, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0289, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11448585987091064, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0302, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.15148931741714478, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0292, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.14388935267925262, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0257, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.14465516805648804, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0272, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.13414210081100464, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0273, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12774589657783508, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.028, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.24874050915241241, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0282, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.1624344289302826, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0294, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.12817884981632233, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0324, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1509743630886078, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0269, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.31345340609550476, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0287, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.1523321270942688, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0259, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.1633133739233017, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0269, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.14278745651245117, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.0253, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.18361203372478485, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0282, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.10244551301002502, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0268, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.125647634267807, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0264, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.12238023430109024, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0264, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.16535502672195435, + "learning_rate": 2.882169202177678e-05, + "loss": 0.0329, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.19624854624271393, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0303, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.18285983800888062, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0266, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.13460052013397217, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.0293, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.1427820920944214, + "learning_rate": 2.881525979398615e-05, + "loss": 0.0244, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.14855900406837463, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0276, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.1433439403772354, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0299, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.1452850103378296, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0271, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.14403793215751648, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0264, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.13841332495212555, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0277, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.1426801085472107, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0296, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.14492706954479218, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0272, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.1315297782421112, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.0289, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.1513436734676361, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0284, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.15784604847431183, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.0273, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.12729156017303467, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0297, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.20208381116390228, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0284, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.1486770659685135, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0271, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.1463368982076645, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.0292, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.12362723797559738, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0252, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.18984590470790863, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0292, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.12962038815021515, + "learning_rate": 2.878773564250086e-05, + "loss": 0.0301, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.132072314620018, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0298, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.1482372134923935, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0267, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.15687556564807892, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0308, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.11656040698289871, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.028, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.16482318937778473, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0256, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.13332504034042358, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.027, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.16453887522220612, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0253, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.16902299225330353, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0283, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.12483953684568405, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0276, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.14033488929271698, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0277, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.18922871351242065, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0258, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.15026360750198364, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0298, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17492659389972687, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0275, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.14793051779270172, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0301, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.17575892806053162, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.027, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.1661873608827591, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0275, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.14529843628406525, + "learning_rate": 2.875990888327069e-05, + "loss": 0.0277, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.11084084957838058, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0278, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1433780938386917, + "learning_rate": 2.875661528053842e-05, + "loss": 0.026, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.13195206224918365, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0298, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.14103750884532928, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0309, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.17067542672157288, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0296, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.14078323543071747, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0288, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.17906904220581055, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0268, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.163010373711586, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0285, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.17636623978614807, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0276, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.18726587295532227, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0297, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.18583180010318756, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.16773444414138794, + "learning_rate": 2.874008460348417e-05, + "loss": 0.0289, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.14579996466636658, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.12655268609523773, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0246, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.15124735236167908, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0297, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.16273140907287598, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0295, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14014655351638794, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0262, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12657561898231506, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0292, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.13813915848731995, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.0287, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.12874144315719604, + "learning_rate": 2.872678492608539e-05, + "loss": 0.026, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.12086840718984604, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0245, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.15290603041648865, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0297, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.13261260092258453, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0263, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.17371073365211487, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0268, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.17300288379192352, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0264, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.16441872715950012, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0279, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.17883789539337158, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.03, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.14519335329532623, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0266, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.18063998222351074, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0275, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.19818824529647827, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0256, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1695982664823532, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0274, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.16905854642391205, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0277, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.1852719783782959, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.03, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.14601513743400574, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0261, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.11720956116914749, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0279, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.14177089929580688, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0268, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.14933669567108154, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0292, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15183576941490173, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0285, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.1255270391702652, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.025, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.1467735320329666, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0258, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.13636574149131775, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0278, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.12642863392829895, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0254, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.09835807979106903, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0284, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.11784205585718155, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0285, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.14453290402889252, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0301, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.19569514691829681, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0264, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.13860125839710236, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.026, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.16651827096939087, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0283, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.13682647049427032, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0309, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.14632774889469147, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0275, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.1401532143354416, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0261, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.12643243372440338, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0268, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1152961477637291, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0259, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.14027607440948486, + "learning_rate": 2.866951776928e-05, + "loss": 0.0306, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.12379784137010574, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0272, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.15595580637454987, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0288, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.12823393940925598, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0263, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1481468379497528, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0283, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.12768998742103577, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0262, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.15771906077861786, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0267, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.18978171050548553, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0295, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.1386374831199646, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0277, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.1583528220653534, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0273, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.09999477118253708, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0253, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.14432361721992493, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0289, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.19001086056232452, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.0295, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.13931159675121307, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0257, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.1621151715517044, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0269, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.15239235758781433, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0275, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12619581818580627, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0273, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.14698320627212524, + "learning_rate": 2.864043385528954e-05, + "loss": 0.0275, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.14121954143047333, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0271, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.16203583776950836, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0265, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.15562862157821655, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0267, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.16438758373260498, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0285, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.13769859075546265, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0274, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.14908984303474426, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0256, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.14057780802249908, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.0261, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.16056638956069946, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.0267, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.12299710512161255, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0257, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.12815117835998535, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0284, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.14666494727134705, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0303, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.15829558670520782, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0243, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1499258577823639, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.029, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.16584458947181702, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0272, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1765419989824295, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0278, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.17066679894924164, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0281, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13183729350566864, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0255, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.12179457396268845, + "learning_rate": 2.86093128283797e-05, + "loss": 0.028, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.10134928673505783, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0275, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16393089294433594, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0278, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1360260546207428, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0297, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.16158054769039154, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0249, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.12037541717290878, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.0286, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.16677190363407135, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0277, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1442420333623886, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0268, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.1756538599729538, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0282, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.11580783128738403, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0282, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.11671322584152222, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0279, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.15163642168045044, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0297, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.1390339583158493, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0282, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.11784064769744873, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0275, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.15270282328128815, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0282, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.09030280262231827, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0281, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.15108083188533783, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0274, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.1403200924396515, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0265, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.14890284836292267, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0303, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.11579708009958267, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0283, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.14590834081172943, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0268, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.1513289511203766, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0264, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.13494084775447845, + "learning_rate": 2.857082133984191e-05, + "loss": 0.026, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.11738834530115128, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.029, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.1497882753610611, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0271, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.17350555956363678, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0277, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.16718043386936188, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0269, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.12658995389938354, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0267, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15880689024925232, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0256, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.1480095237493515, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0263, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.13779640197753906, + "learning_rate": 2.85567006866004e-05, + "loss": 0.0282, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1088612824678421, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0245, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.17420272529125214, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0267, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.12365500628948212, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.025, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.13261641561985016, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0256, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.10485266894102097, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0253, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14070947468280792, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0284, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.1179763451218605, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0241, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13687312602996826, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0284, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.17160922288894653, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0263, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.17487400770187378, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0275, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.11352817714214325, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1871306300163269, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0277, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.13434743881225586, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0278, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.16361933946609497, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0257, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.19457586109638214, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0277, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.16905367374420166, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0323, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.14845560491085052, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0248, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.12934044003486633, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.026, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.11046823114156723, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0267, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.14174290001392365, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0255, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.15194296836853027, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.0257, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.1532754749059677, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0269, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19864243268966675, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0288, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.16765300929546356, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0271, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.15602844953536987, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0288, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.1571321040391922, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0264, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1625281423330307, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0268, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.12808986008167267, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0307, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.12471310794353485, + "learning_rate": 2.850496147627652e-05, + "loss": 0.027, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.1521507054567337, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0289, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1389119178056717, + "learning_rate": 2.850136142292186e-05, + "loss": 0.024, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1338133066892624, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13214652240276337, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.0274, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.17002162337303162, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0277, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.1433296948671341, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0267, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.19547788798809052, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0287, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.1916888952255249, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0296, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.17188303172588348, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0262, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.14335697889328003, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.0269, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14298376441001892, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0278, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.12892583012580872, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0278, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.16246432065963745, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0255, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.1237327978014946, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0252, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.12249717861413956, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0238, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.16828545928001404, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0262, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.12834671139717102, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0287, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.11583935469388962, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.029, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.10479022562503815, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0244, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.18407097458839417, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0276, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.12898766994476318, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0277, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.13034385442733765, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0241, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.13197654485702515, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0269, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.14327986538410187, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0281, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.15228267014026642, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0269, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.11858481913805008, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0261, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.16075590252876282, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0284, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.1527116447687149, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0283, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.1280883252620697, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0294, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1348433941602707, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0251, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.15039700269699097, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0272, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.16682399809360504, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.0277, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.11187390983104706, + "learning_rate": 2.844503681416359e-05, + "loss": 0.0298, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.1675758808851242, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0267, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.16022244095802307, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0301, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.16192372143268585, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0277, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.13092738389968872, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.12416383624076843, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.028, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.12514744699001312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0252, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.1705007702112198, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.0274, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.14251261949539185, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0264, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.147093266248703, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0279, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.12747326493263245, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0265, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.14222867786884308, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0276, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.12574070692062378, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0262, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.12812504172325134, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0225, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.13585571944713593, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0267, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.12168630957603455, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0236, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.12048719823360443, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0225, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.15165774524211884, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0267, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.15395723283290863, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.026, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.14715734124183655, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0259, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.11370360106229782, + "learning_rate": 2.840817735753433e-05, + "loss": 0.0284, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1264869123697281, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.028, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.13726982474327087, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0268, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1450515240430832, + "learning_rate": 2.840261328154438e-05, + "loss": 0.0251, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.1328781247138977, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.025, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.15474216639995575, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.0292, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.1338219791650772, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0279, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.16368664801120758, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.027, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.18125605583190918, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.0252, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.127846822142601, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0252, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.17743080854415894, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0275, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.16253551840782166, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0287, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.14212867617607117, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0274, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.1380675882101059, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0263, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.12773451209068298, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.0257, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.1556098759174347, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0264, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.16690294444561005, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0296, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.11710768938064575, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0252, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.17441801726818085, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0318, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.10790088772773743, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0251, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.16537238657474518, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0266, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.1414048671722412, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0276, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.1096765398979187, + "learning_rate": 2.836716156259365e-05, + "loss": 0.0258, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.14450953900814056, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0278, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15733887255191803, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0278, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.15253902971744537, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0275, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.1410975307226181, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0249, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1242721900343895, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0251, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.15358713269233704, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0287, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.15856830775737762, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.029, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.1777517944574356, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0254, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.10823221504688263, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0262, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.144780695438385, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0286, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12900282442569733, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0243, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.15370796620845795, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0238, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.15801627933979034, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0259, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.19058606028556824, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0257, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14663439989089966, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0281, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.10609344393014908, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.0261, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.11403398960828781, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0291, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.11631747335195541, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0243, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.1325336992740631, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0266, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.11245301365852356, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.10258981585502625, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0252, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.15847361087799072, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0258, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.15267148613929749, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0246, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.15297843515872955, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0248, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.1767774075269699, + "learning_rate": 2.831995639121568e-05, + "loss": 0.0292, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.12696890532970428, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.025, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.17774799466133118, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0267, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.15603452920913696, + "learning_rate": 2.83142492407012e-05, + "loss": 0.0285, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.12731599807739258, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0272, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.13816066086292267, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0265, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.14216956496238708, + "learning_rate": 2.830853299002291e-05, + "loss": 0.026, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1427476406097412, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0263, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.14667609333992004, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0282, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.11785806715488434, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0276, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.125101238489151, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0255, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.10735481232404709, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0248, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.14256443083286285, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0286, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12700040638446808, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0275, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.11069332808256149, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0283, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.1438131332397461, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.0269, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.1901932805776596, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0274, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.17823027074337006, + "learning_rate": 2.828557706390495e-05, + "loss": 0.029, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.13579529523849487, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0267, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.1431250274181366, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0279, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.11313948035240173, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0274, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1669740527868271, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0288, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.12642791867256165, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0267, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.15632158517837524, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0256, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.1404937356710434, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0263, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.13466882705688477, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0249, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.13014933466911316, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0255, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.18688003718852997, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0267, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.14594615995883942, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0253, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.1708635836839676, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0244, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.16193392872810364, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0265, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.13249778747558594, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0286, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.13875705003738403, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0262, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.19137892127037048, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0246, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.15596376359462738, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0248, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.1577422320842743, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0247, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.14322742819786072, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0263, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.13651713728904724, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0261, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11288725584745407, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0252, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.09636299312114716, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0267, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.11729153245687485, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0275, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.14827747642993927, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0265, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.1446153074502945, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0257, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.1329014152288437, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0247, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.12726181745529175, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0259, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.13508957624435425, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0258, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.14538758993148804, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0264, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.10585593432188034, + "learning_rate": 2.822755216255228e-05, + "loss": 0.0276, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.17435809969902039, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0281, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.1504661589860916, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0259, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.1678706258535385, + "learning_rate": 2.8221699881734e-05, + "loss": 0.029, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.13508610427379608, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0261, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12610375881195068, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0237, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.12272591143846512, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0231, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.1211056336760521, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0252, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.13204888999462128, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0276, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.11982017755508423, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0245, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.17140068113803864, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0236, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.12660211324691772, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0264, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1636001467704773, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0266, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.16939589381217957, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0276, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.15193505585193634, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0291, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16562964022159576, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0283, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.13603557646274567, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.024, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.14359496533870697, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0257, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.14077764749526978, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0247, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.14604462683200836, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0277, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.1240646094083786, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0262, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13144497573375702, + "learning_rate": 2.818639656220465e-05, + "loss": 0.0255, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.1384861171245575, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0227, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.16475403308868408, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0269, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.14310070872306824, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0258, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1498526632785797, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0263, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.13084106147289276, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0268, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.12004188448190689, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0252, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.14093606173992157, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0262, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.11016195267438889, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.0249, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.11101678013801575, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0246, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.1335095763206482, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0274, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1479913741350174, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0237, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.11868918687105179, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0231, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12799321115016937, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.027, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.17011503875255585, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0271, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.15915420651435852, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0274, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.11501206457614899, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0254, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.1457805633544922, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0274, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.1501278430223465, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0264, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.13190113008022308, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0272, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.14392758905887604, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0235, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.1375834047794342, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0241, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.14287824928760529, + "learning_rate": 2.814280752477756e-05, + "loss": 0.0269, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12234462797641754, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0259, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.1336476355791092, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0242, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.11315035820007324, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0245, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.14132073521614075, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0293, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.12053677439689636, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.0266, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.10896297544240952, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0268, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.11100424081087112, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.026, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.11256317049264908, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0237, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.14322060346603394, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0255, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.1655588448047638, + "learning_rate": 2.812283453196085e-05, + "loss": 0.026, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.14714404940605164, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0261, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.12317050993442535, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0247, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12872155010700226, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0258, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.1666097790002823, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0261, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.13425199687480927, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.16131597757339478, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0295, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.13762447237968445, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0242, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.12883082032203674, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0274, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.15029405057430267, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0269, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.13236507773399353, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0303, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.12054237723350525, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0218, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.1471109539270401, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0242, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.11926044523715973, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0275, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.13649292290210724, + "learning_rate": 2.809470494867301e-05, + "loss": 0.0231, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.15434928238391876, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0283, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.11536315083503723, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0271, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.12312797456979752, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0249, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.12750110030174255, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0235, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.1568569540977478, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0277, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.11667264997959137, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0258, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.14598897099494934, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0266, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.21723008155822754, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0258, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.14233075082302094, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0244, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.14836373925209045, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0265, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.146131694316864, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0258, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.16575932502746582, + "learning_rate": 2.807043872632495e-05, + "loss": 0.025, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.15142853558063507, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0245, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.11762700974941254, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0253, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.09714391082525253, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0233, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.1586424708366394, + "learning_rate": 2.806231820878714e-05, + "loss": 0.023, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.14317330718040466, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.0271, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.10633813589811325, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.0244, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.19832657277584076, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0249, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.14428019523620605, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0283, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.18295790255069733, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.0257, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.16204482316970825, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0285, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.14742101728916168, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0239, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1435142159461975, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0253, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.2000540792942047, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.16311310231685638, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0253, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.1238052025437355, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0249, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.16693037748336792, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0269, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.12382039427757263, + "learning_rate": 2.803581696238626e-05, + "loss": 0.025, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.13866499066352844, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0259, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.12111294269561768, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0258, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.13717366755008698, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0266, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.1119573786854744, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0264, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.1390918791294098, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.12881262600421906, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0251, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.1354016363620758, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0256, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.1285468488931656, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0315, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.17634150385856628, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0254, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.15077067911624908, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.0268, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16464342176914215, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0265, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.3138365149497986, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0255, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.15954458713531494, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0272, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15878041088581085, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0267, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.17743942141532898, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0273, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.16661205887794495, + "learning_rate": 2.800297040373295e-05, + "loss": 0.0268, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.13536596298217773, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.0241, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.13258172571659088, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0266, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.1524200588464737, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0276, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.12779775261878967, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0245, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1008654311299324, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0222, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.14413924515247345, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0297, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.14558148384094238, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0252, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.15591812133789062, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0248, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.12246204912662506, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0276, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.13106316328048706, + "learning_rate": 2.7982312890369e-05, + "loss": 0.0265, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.12663032114505768, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0234, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.1167084351181984, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15868185460567474, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0251, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13669976592063904, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0291, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.15100552141666412, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0237, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14283764362335205, + "learning_rate": 2.796987104810136e-05, + "loss": 0.024, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.13055892288684845, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0255, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.16733981668949127, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0244, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.1716628074645996, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0257, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.11838851124048233, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0288, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.14557431638240814, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0238, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.14981132745742798, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0254, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.11149631440639496, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0256, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.20500904321670532, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0255, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.12128293514251709, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0254, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.12934274971485138, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0254, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.1355874240398407, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0267, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13485173881053925, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0259, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.1481594294309616, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0251, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.15145331621170044, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0243, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.16746139526367188, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0284, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.10760390013456345, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0268, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.1460299789905548, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0264, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.15936219692230225, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0232, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.1373136043548584, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.0255, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.14508306980133057, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0243, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.153093621134758, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0262, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.15373852849006653, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0238, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.14033931493759155, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0267, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.1424216777086258, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0252, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.11820772290229797, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0253, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.13179944455623627, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0243, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.18073157966136932, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0256, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.12908786535263062, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0254, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.18455258011817932, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0263, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.14844198524951935, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0248, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.14291873574256897, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0268, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.105900838971138, + "learning_rate": 2.79029165248073e-05, + "loss": 0.027, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.14083242416381836, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0244, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.14579042792320251, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0239, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.13705924153327942, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0254, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.17981873452663422, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0225, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.13812783360481262, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0278, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12511365115642548, + "learning_rate": 2.789025068651819e-05, + "loss": 0.027, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.13009224832057953, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0245, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.145152747631073, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0281, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.1607820838689804, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0254, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.09400778263807297, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0258, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.1608203500509262, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0263, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.14968201518058777, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0247, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.1333402544260025, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0272, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.17871612310409546, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0256, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.10688412934541702, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0264, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.154439315199852, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0259, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.1270124316215515, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0279, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.15214042365550995, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0259, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.14094167947769165, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0261, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.11419129371643066, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.025, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.1447947919368744, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0258, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.12327348440885544, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0246, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.11033894866704941, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0259, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.128106951713562, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0247, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.13206088542938232, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0254, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.13664662837982178, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0274, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1349778026342392, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0262, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.10689893364906311, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0262, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.16241611540317535, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0269, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14955222606658936, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0251, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.13519948720932007, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0231, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.15076202154159546, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0272, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.15367662906646729, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0255, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.15819156169891357, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0242, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.12844862043857574, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0259, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.13392682373523712, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0245, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.15134917199611664, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.0287, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.14139734208583832, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0254, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.1562367081642151, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0246, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.1298898160457611, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0252, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13048380613327026, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0252, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.12535928189754486, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0271, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.19301816821098328, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0245, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.12262707948684692, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0262, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1394580751657486, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0272, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.21376213431358337, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0243, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.15442071855068207, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0285, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.13063938915729523, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0277, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.15023048222064972, + "learning_rate": 2.779844963973707e-05, + "loss": 0.025, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.17523249983787537, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0239, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.1312434822320938, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0234, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13352124392986298, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0249, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.12799884378910065, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0257, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.12704412639141083, + "learning_rate": 2.778765821684581e-05, + "loss": 0.025, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.11119242757558823, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0248, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.1564178764820099, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0282, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.13500694930553436, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0279, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.1542324423789978, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0258, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13888850808143616, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0281, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.14011257886886597, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.14887000620365143, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0254, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.15670792758464813, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0289, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.1288663148880005, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0242, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.13191169500350952, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.027, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.1320677250623703, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0239, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.13418279588222504, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0234, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.11019188910722733, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0232, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.13963419198989868, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0235, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.1163327619433403, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.028, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.13508521020412445, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0275, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.11098913848400116, + "learning_rate": 2.775078589916535e-05, + "loss": 0.025, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.16348809003829956, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0259, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.12986604869365692, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0246, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1302310824394226, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0259, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.11797303706407547, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.0252, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.1427793651819229, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0233, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.13411074876785278, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.0257, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.1528121531009674, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.11762527376413345, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0238, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.13639862835407257, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0257, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.1356336623430252, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0269, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.16879825294017792, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.13352762162685394, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0236, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.20271515846252441, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0296, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.11791368573904037, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0241, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14179861545562744, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.0256, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12654612958431244, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0243, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12589605152606964, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0222, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.12714968621730804, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0278, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.13410630822181702, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0248, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.12528881430625916, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0258, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.11806205660104752, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0308, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.14317870140075684, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0244, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14099767804145813, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0228, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.15186408162117004, + "learning_rate": 2.769825463131276e-05, + "loss": 0.0252, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.18532416224479675, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0262, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.1445159614086151, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0265, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14576950669288635, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0253, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1777651309967041, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0229, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.16440069675445557, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0234, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.12858232855796814, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0259, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.14329349994659424, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0254, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.1265982836484909, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0235, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13783425092697144, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0228, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.1327633559703827, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0248, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14867408573627472, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0257, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.1556752771139145, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0239, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.16407394409179688, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0261, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.14142248034477234, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0265, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.11142456531524658, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0221, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.127807155251503, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0265, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.14716863632202148, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0241, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.1455392986536026, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.024, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1665087193250656, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0253, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.16937066614627838, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0292, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.13188236951828003, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0254, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.1010027602314949, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.0236, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.15787169337272644, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.0279, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.169437438249588, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0246, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.22705122828483582, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0227, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.17403557896614075, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.027, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.1287415474653244, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0253, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.13813741505146027, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0238, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.11655540764331818, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.13267700374126434, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.025, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.10447681695222855, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0239, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.11865647882223129, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.025, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.13245069980621338, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0256, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.11377252638339996, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0266, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.11403302848339081, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0252, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.1518944650888443, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0286, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.13505733013153076, + "learning_rate": 2.761618148370765e-05, + "loss": 0.0253, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.12230793386697769, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0247, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.1600245237350464, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0267, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.13100600242614746, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.11071992665529251, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0254, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.16208025813102722, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.0287, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.1652526706457138, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.027, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14042513072490692, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.1674395203590393, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0257, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.138246089220047, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0272, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13758982717990875, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0227, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.15503156185150146, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0259, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.15050673484802246, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0279, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.13952329754829407, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.024, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.14903564751148224, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0242, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.09984514862298965, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.09952808916568756, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0233, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.14514309167861938, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0255, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.12824438512325287, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0232, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.15507018566131592, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0251, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.12279994785785675, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0225, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.13601353764533997, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0249, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.20022229850292206, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.0263, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.1407146006822586, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0226, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.1350381076335907, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0248, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.13220156729221344, + "learning_rate": 2.755998359318217e-05, + "loss": 0.0251, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.11746657639741898, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0256, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.10933379083871841, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0235, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.1263892501592636, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0244, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.14062181115150452, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.0255, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.1430615335702896, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0254, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.19069550931453705, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0254, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.16568611562252045, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0252, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.15911205112934113, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.0265, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.17054520547389984, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0265, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.15547682344913483, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0237, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.12778308987617493, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0242, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.1590787172317505, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0232, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13945519924163818, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.023, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.14507925510406494, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0247, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14541085064411163, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0245, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.17666012048721313, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0222, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.16692468523979187, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0218, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.1295796036720276, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0242, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.12527599930763245, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0273, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.13334719836711884, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0253, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.1175302267074585, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0251, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.14461147785186768, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0261, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.15658612549304962, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0236, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.15802229940891266, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0236, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.14738211035728455, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0252, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.1639637053012848, + "learning_rate": 2.750090542643057e-05, + "loss": 0.027, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1573086827993393, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0264, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.13114139437675476, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0238, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.1112227514386177, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.027, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.12209689617156982, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.13741743564605713, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0239, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14444704353809357, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0263, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.13822096586227417, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0246, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.14305464923381805, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0237, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13539855182170868, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0264, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.13791607320308685, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0234, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.205885648727417, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0234, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.1034194752573967, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0214, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.15530072152614594, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0258, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.1535874903202057, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0226, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.19778263568878174, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0238, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.175798699259758, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0255, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.1803887039422989, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0238, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.1906212419271469, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0253, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.13742655515670776, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.026, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.12772122025489807, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0261, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13805702328681946, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0259, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.14081542193889618, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0241, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.10023088753223419, + "learning_rate": 2.744810891043728e-05, + "loss": 0.0246, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.10111884027719498, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0215, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.11498598754405975, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0217, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10912947356700897, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.0257, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.13553178310394287, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0266, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.18210133910179138, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0234, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.11901862919330597, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0248, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.14622537791728973, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0234, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.1641179323196411, + "learning_rate": 2.742962757296782e-05, + "loss": 0.026, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.14369463920593262, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.0236, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.14231407642364502, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0213, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.10929716378450394, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0216, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.13811182975769043, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0241, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.13219693303108215, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0219, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.13322590291500092, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0259, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.14716853201389313, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0255, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.11750922352075577, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0272, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.1430707424879074, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1722290813922882, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.025, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.15502606332302094, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.0231, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.09124620258808136, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0246, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.16296057403087616, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0277, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.10881830751895905, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0222, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.11321897804737091, + "learning_rate": 2.739481230505444e-05, + "loss": 0.026, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.1362120658159256, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.03, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.14329490065574646, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0252, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.4512016475200653, + "learning_rate": 2.738782381719405e-05, + "loss": 0.0238, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12559202313423157, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0263, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.12616541981697083, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0248, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.1516106277704239, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0234, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.12785211205482483, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0264, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16447393596172333, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.026, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.1648014783859253, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0264, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.14414788782596588, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0249, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.10982059687376022, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0254, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.12677313387393951, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0263, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.12355965375900269, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0223, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.15190085768699646, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0246, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.13910633325576782, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0231, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.11962535977363586, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0259, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.1566285789012909, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0256, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.12282316386699677, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0253, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.16674686968326569, + "learning_rate": 2.735040900116229e-05, + "loss": 0.025, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.14292193949222565, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0277, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.1504344791173935, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0268, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.16317704319953918, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0229, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.20625433325767517, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0242, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.1895458698272705, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.024, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.1987384408712387, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.0258, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.18668580055236816, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.1807073950767517, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0246, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.11927719414234161, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.028, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.11955022811889648, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0259, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.15261328220367432, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0268, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.15620580315589905, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0254, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.12000948935747147, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0234, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.1188279539346695, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.0248, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.13279424607753754, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0241, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.13379216194152832, + "learning_rate": 2.731275407491117e-05, + "loss": 0.026, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.12054166197776794, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0215, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.15941590070724487, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0279, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.18678002059459686, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0263, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.21807490289211273, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.0239, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.12243734300136566, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.028, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.14628900587558746, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0248, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.17471152544021606, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0257, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.11189062893390656, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0242, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.16700562834739685, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0252, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.1213584691286087, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0251, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.14496786892414093, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0258, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.17899028956890106, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0248, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13594108819961548, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0246, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.1365566849708557, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.021, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.10815171152353287, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0227, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.11390877515077591, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0209, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.1541927456855774, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0213, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.1342182159423828, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0251, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.1300208568572998, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0249, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14917100965976715, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0274, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.14349427819252014, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.0269, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10514543205499649, + "learning_rate": 2.72605878418951e-05, + "loss": 0.024, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.12946246564388275, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0234, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.1120578870177269, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.0246, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.14252731204032898, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.0287, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.0984109416604042, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0234, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.13058412075042725, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0235, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.1468561738729477, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0263, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.1297929584980011, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0224, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.15071524679660797, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0254, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.15211531519889832, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0239, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.13758648931980133, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0254, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.13972750306129456, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0239, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1762881577014923, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0235, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.13277511298656464, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.025, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.11352386325597763, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0225, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.1340995877981186, + "learning_rate": 2.722476145603374e-05, + "loss": 0.0232, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.16372360289096832, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0224, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.130090594291687, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0231, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.10481512546539307, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0229, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.1453929841518402, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0255, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.133439838886261, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0248, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.1380738615989685, + "learning_rate": 2.721037237880308e-05, + "loss": 0.023, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.1386265754699707, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0264, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.14298051595687866, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.0247, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.11756420135498047, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0222, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1197751984000206, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0216, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.2194000482559204, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0243, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.14157775044441223, + "learning_rate": 2.719594991888059e-05, + "loss": 0.023, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.14389759302139282, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0228, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.12442908436059952, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0267, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.1188371554017067, + "learning_rate": 2.718872618272836e-05, + "loss": 0.027, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11137005686759949, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0237, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.10907725989818573, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0251, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.150213360786438, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0229, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.2098330855369568, + "learning_rate": 2.717908157623383e-05, + "loss": 0.0245, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.12558938562870026, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0257, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.13659170269966125, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0247, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.12037219852209091, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0228, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.13085514307022095, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0252, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.13916544616222382, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0251, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.13098886609077454, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.021, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.12777045369148254, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0258, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.10204357653856277, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0249, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.11189272999763489, + "learning_rate": 2.715732712163776e-05, + "loss": 0.0232, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.1412421613931656, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0255, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.09470678865909576, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0217, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.09736761450767517, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0233, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12988215684890747, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.1080017015337944, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0263, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.12020932137966156, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0238, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.1058862954378128, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0247, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11686211824417114, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.024, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.11078432947397232, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0226, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.13394472002983093, + "learning_rate": 2.713306780076029e-05, + "loss": 0.0237, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.11299733072519302, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.025, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.1196199283003807, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0251, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.16477037966251373, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0241, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.1175335943698883, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.11024156212806702, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0237, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.12922869622707367, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0278, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.1429518610239029, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0254, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.14210145175457, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0225, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.12270870059728622, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0257, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.12436466664075851, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0258, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.10704488307237625, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0224, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.15890415012836456, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0252, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.12372975796461105, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0225, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.11652565747499466, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.028, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.1483478546142578, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0255, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.13188686966896057, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0238, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.18725869059562683, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0244, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.15949182212352753, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0225, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.14752954244613647, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0229, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.14210951328277588, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0233, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.1388787031173706, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.0253, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.11010867357254028, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0227, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.13336122035980225, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0227, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.10118173807859421, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.0265, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.1434611827135086, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0263, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.14133326709270477, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.026, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.15359432995319366, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0236, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.12927602231502533, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.022, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.10330940783023834, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0236, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.1264553815126419, + "learning_rate": 2.705973772304842e-05, + "loss": 0.025, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.15007789433002472, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0267, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.11090201884508133, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0257, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.12769529223442078, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0252, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.0977141410112381, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0244, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.14598220586776733, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0254, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.11231301724910736, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0235, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.09610529243946075, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0262, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.1549275815486908, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0234, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.11284032464027405, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.0217, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.14465086162090302, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.0235, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.13792403042316437, + "learning_rate": 2.703264323632037e-05, + "loss": 0.024, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.1438492238521576, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0239, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.12613703310489655, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0244, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.14072397351264954, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0225, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.1420842856168747, + "learning_rate": 2.702276326859174e-05, + "loss": 0.026, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.17447218298912048, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0243, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.1708984673023224, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0259, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.1555069535970688, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.0252, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.12473946064710617, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0227, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.15750949084758759, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0248, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.14009647071361542, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0267, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.1416911780834198, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0229, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.11944608390331268, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0235, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.126204252243042, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0268, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.1186814159154892, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0242, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.12699876725673676, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0239, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.10452742129564285, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0221, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.1422647386789322, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0247, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.11601094901561737, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0232, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12339281290769577, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0223, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1883407086133957, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0243, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.1730756312608719, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0294, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.10429747402667999, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0194, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.1358909010887146, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0217, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.14417894184589386, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0265, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.12299799919128418, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0236, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.1361788511276245, + "learning_rate": 2.696816263672279e-05, + "loss": 0.024, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.14492252469062805, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0241, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.1209225058555603, + "learning_rate": 2.696317709902802e-05, + "loss": 0.0239, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.11048539727926254, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.12423776835203171, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0243, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.1254356950521469, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.12912537157535553, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0254, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.10864657163619995, + "learning_rate": 2.695069735893089e-05, + "loss": 0.022, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.15519478917121887, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0253, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.15276901423931122, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0231, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.1401076465845108, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0271, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.14237190783023834, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.024, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1186559721827507, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0238, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.1254940629005432, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0233, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.1337885558605194, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0255, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12958411872386932, + "learning_rate": 2.69306825906724e-05, + "loss": 0.025, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.11610005795955658, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0238, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.13515442609786987, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0239, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1492868959903717, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0224, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.1293300986289978, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.021, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12354710698127747, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.13226614892482758, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.0247, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.1368253529071808, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0255, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.13044875860214233, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0228, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14107753336429596, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0221, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.13521252572536469, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0236, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.13718804717063904, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0217, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.15192179381847382, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.0247, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.13718710839748383, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0244, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11781664192676544, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0239, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.14431379735469818, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0248, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.15059512853622437, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0231, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.12076479196548462, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0242, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.19006308913230896, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0238, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.140592560172081, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0233, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.1528412401676178, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0229, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.16418038308620453, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0242, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.12185576558113098, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0254, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11594245582818985, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0248, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1216709092259407, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0234, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.12024594098329544, + "learning_rate": 2.686776313743295e-05, + "loss": 0.024, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.1321716010570526, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.0255, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.13785192370414734, + "learning_rate": 2.686270520700719e-05, + "loss": 0.025, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.12321551889181137, + "learning_rate": 2.6860174890357e-05, + "loss": 0.0233, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14606279134750366, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0249, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.17512187361717224, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0272, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.14974310994148254, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.0243, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.11758870631456375, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0241, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.11157143115997314, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0232, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.10899091511964798, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0231, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.1136092096567154, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0224, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.10306857526302338, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0211, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13241226971149445, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0222, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.11886361986398697, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0244, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.11962216347455978, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0237, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.14516247808933258, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.14592237770557404, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0253, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1278424859046936, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0233, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.12451537698507309, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0218, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.11746378242969513, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0266, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.14618739485740662, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0226, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.1512313336133957, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0233, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.14035750925540924, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0264, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.13232699036598206, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.023, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.14616237580776215, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.0233, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.12882372736930847, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0265, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.13778620958328247, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.024, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.15017054975032806, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0239, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.14781621098518372, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0215, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1670294851064682, + "learning_rate": 2.679407107451536e-05, + "loss": 0.022, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.1112075075507164, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0232, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.12219138443470001, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0202, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12922540307044983, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.11967732757329941, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0226, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.1377563774585724, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0224, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.14771930873394012, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0234, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.14800170063972473, + "learning_rate": 2.677617030053916e-05, + "loss": 0.0251, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.1233784481883049, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0232, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.16905057430267334, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0214, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.13751909136772156, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0248, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.13289029896259308, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0216, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.10788674652576447, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0216, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.13367260992527008, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1270209699869156, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0239, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.140412375330925, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0218, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.17788565158843994, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0253, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.13695330917835236, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.023, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.1250872015953064, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0232, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.11347977817058563, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0243, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.15546919405460358, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0233, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.13752058148384094, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.10431966185569763, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0215, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.17878347635269165, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0262, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.13218404352664948, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0241, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.12244930118322372, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.0247, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.11582386493682861, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0219, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.1160157322883606, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0203, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.10731764137744904, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0234, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.11152671277523041, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0218, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.12760375440120697, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0218, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.13706916570663452, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0219, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.14939993619918823, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0208, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.1169784739613533, + "learning_rate": 2.67092985767871e-05, + "loss": 0.0217, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.1569128930568695, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12233320623636246, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0251, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13152818381786346, + "learning_rate": 2.670154386126402e-05, + "loss": 0.0231, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.1529632955789566, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0246, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.11001774668693542, + "learning_rate": 2.669636960731296e-05, + "loss": 0.0223, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.1473267376422882, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.025, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.12853027880191803, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0246, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.15698745846748352, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0209, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12175270169973373, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0229, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.15634675323963165, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.0224, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.16060872375965118, + "learning_rate": 2.668082553346281e-05, + "loss": 0.0254, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.13773131370544434, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0224, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.14258436858654022, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0216, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.13032476603984833, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0215, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.17300327122211456, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0221, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.13187018036842346, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0226, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.14065369963645935, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0232, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.11484711617231369, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0263, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1380661576986313, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0209, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.17800134420394897, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0244, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.14836716651916504, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0277, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14563506841659546, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.023, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.11463280767202377, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.026, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.13092686235904694, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0222, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.12265385687351227, + "learning_rate": 2.664443191002912e-05, + "loss": 0.025, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.14121955633163452, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0242, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.12938837707042694, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0222, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.1922816038131714, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.0228, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.12596914172172546, + "learning_rate": 2.663400187275054e-05, + "loss": 0.022, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.14095240831375122, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0244, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.1080891340970993, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0228, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.13916714489459991, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.0239, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.1489475965499878, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0251, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.11781706660985947, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0207, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.11544635146856308, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0232, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.14827974140644073, + "learning_rate": 2.661571529941175e-05, + "loss": 0.0236, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.12821446359157562, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.0186, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.1303764134645462, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0234, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.1049366369843483, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0232, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.15925322473049164, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0275, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.13151033222675323, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0213, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.09425143152475357, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0215, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.15108847618103027, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0228, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.1374080330133438, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0231, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.16061817109584808, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0253, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.1707434207201004, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0259, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.12051897495985031, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0217, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.09191743284463882, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0224, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.12547235190868378, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0245, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14491257071495056, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0228, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.11896242201328278, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0239, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.14253611862659454, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0233, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.12803201377391815, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.0226, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.12893305718898773, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0231, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.16127154231071472, + "learning_rate": 2.656586262343925e-05, + "loss": 0.0234, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.13307608664035797, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0258, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.11211924254894257, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.0219, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.1672789305448532, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0248, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.12878142297267914, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0231, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.11429315060377121, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0233, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.13469499349594116, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0254, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.12753085792064667, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0258, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.11875704675912857, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0234, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.13718342781066895, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0223, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.10762278735637665, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.021, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.11651597917079926, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0222, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.12953802943229675, + "learning_rate": 2.653421326308085e-05, + "loss": 0.0235, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.14820824563503265, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.023, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.16258545219898224, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0266, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.14097236096858978, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0224, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.1039188802242279, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0216, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15014079213142395, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0246, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12850438058376312, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.023, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.12245986610651016, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0209, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.15426407754421234, + "learning_rate": 2.651304359235998e-05, + "loss": 0.025, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.13411583006381989, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0229, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.13705122470855713, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0254, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.13120216131210327, + "learning_rate": 2.650509053370553e-05, + "loss": 0.022, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14599482715129852, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0249, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.11628340184688568, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0215, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.13643144071102142, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0228, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.12846967577934265, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0214, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.13181492686271667, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0245, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1277213841676712, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0244, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12288910150527954, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0226, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.14789454638957977, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0215, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.11457780748605728, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0248, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.12511487305164337, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0206, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.12255935370922089, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0245, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.13214053213596344, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0238, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.1601077914237976, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.027, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.1244364008307457, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0212, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.11869437992572784, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0239, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.14327670633792877, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0227, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.12729141116142273, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0248, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.10953658074140549, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0205, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.20326738059520721, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0225, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.15614672005176544, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0237, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.11863692849874496, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0208, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.10683806240558624, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0224, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.1529647260904312, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.024, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.18428578972816467, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0232, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.19935141503810883, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.0223, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.14675194025039673, + "learning_rate": 2.643583521046136e-05, + "loss": 0.0216, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.21206484735012054, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0243, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.14521193504333496, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0214, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.11565426737070084, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0236, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.1087750792503357, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0215, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.11942578107118607, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0229, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.138235405087471, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0219, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.1079302579164505, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0222, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.15928085148334503, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0225, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.10920153558254242, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.024, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.12247785925865173, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0205, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.10454121232032776, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0238, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.1726839691400528, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0211, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1617811918258667, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0225, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.15452679991722107, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0214, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1573309749364853, + "learning_rate": 2.639561293166059e-05, + "loss": 0.022, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.13596323132514954, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0234, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.09188699722290039, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0219, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.12158972024917603, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0202, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.16615170240402222, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.0221, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.1163380891084671, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0255, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.17704762518405914, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0243, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.1591942012310028, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.021, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.14697043597698212, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0241, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1234351247549057, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0261, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.1379440575838089, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0266, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1025586649775505, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.022, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.11894206702709198, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0226, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1431341916322708, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0241, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.10291697829961777, + "learning_rate": 2.635789644394928e-05, + "loss": 0.022, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.1622016727924347, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0229, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.11867184937000275, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0262, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.18177814781665802, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0237, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.15159954130649567, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0218, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.11481288820505142, + "learning_rate": 2.634438526911474e-05, + "loss": 0.024, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.10989677160978317, + "learning_rate": 2.634168044873305e-05, + "loss": 0.024, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.12370797991752625, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0235, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1390318125486374, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0231, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.14030151069164276, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.0207, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11659140884876251, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.021, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.16727864742279053, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0242, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.11524021625518799, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0265, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.10776443779468536, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0224, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.16366012394428253, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0264, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.1102370023727417, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0245, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11543556302785873, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0216, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.11101235449314117, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0208, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.1257167011499405, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.024, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.1063276082277298, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0241, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.14303457736968994, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0242, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.12862786650657654, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0244, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15487778186798096, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.14244592189788818, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0211, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1066693440079689, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0199, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11721482872962952, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0214, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.13803230226039886, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.0227, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.11225558817386627, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.028, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.11660541594028473, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0239, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12551093101501465, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0221, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.1305139809846878, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.1281510442495346, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0235, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.14735114574432373, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0213, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.10699455440044403, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0231, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.16579580307006836, + "learning_rate": 2.626559653214893e-05, + "loss": 0.0213, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.09389153122901917, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0218, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.17565062642097473, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0245, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.1368194818496704, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0214, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.10599412024021149, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0221, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.12645669281482697, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0233, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.12325828522443771, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.0228, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.11353788524866104, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.10889376699924469, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0227, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.12935040891170502, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0214, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.1422247439622879, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0219, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.16723784804344177, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0239, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.09775881469249725, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.0229, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.13455381989479065, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0253, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.13700921833515167, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0239, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12252738326787949, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0231, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.15983854234218597, + "learning_rate": 2.622181856488204e-05, + "loss": 0.0223, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.18716242909431458, + "learning_rate": 2.621907518918971e-05, + "loss": 0.025, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.16380536556243896, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0229, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.1524866819381714, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0221, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.12281060218811035, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0213, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.09793791174888611, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0249, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.12271476536989212, + "learning_rate": 2.620534553467634e-05, + "loss": 0.0221, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.12616175413131714, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0236, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.10274055600166321, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0205, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.16873067617416382, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.027, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.10157257318496704, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0236, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.11863094568252563, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0225, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.09981630742549896, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0233, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11045375466346741, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0226, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.13111108541488647, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0232, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.12417030334472656, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0226, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.12029916048049927, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0217, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.10766672343015671, + "learning_rate": 2.617506544598029e-05, + "loss": 0.025, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.12039721757173538, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0228, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.11044551432132721, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0223, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.12463348358869553, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.0232, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.14382848143577576, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0238, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.13677646219730377, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0239, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.11536712944507599, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0224, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.12461430579423904, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0253, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.16901540756225586, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0242, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.12255802005529404, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0212, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.14185196161270142, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.0218, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.1292722374200821, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0249, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12001507729291916, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0222, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.12286248803138733, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0211, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.12487491965293884, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0231, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13417202234268188, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0257, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.11690660566091537, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0219, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15474916994571686, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0223, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.1720103919506073, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0239, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.16786889731884003, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0216, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.2279045283794403, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.0254, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13872499763965607, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0238, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.1881643831729889, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0211, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.1372661143541336, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0237, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.10514581203460693, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0217, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.11103228479623795, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0248, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1349911093711853, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0237, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.12774604558944702, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0245, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.15125125646591187, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0233, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.1190832257270813, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0248, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.1088283434510231, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0187, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.12167634069919586, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.023, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.15135347843170166, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0226, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.12857669591903687, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.0195, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.13588359951972961, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0234, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.14902308583259583, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0235, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.15095068514347076, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0209, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.1116727739572525, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0255, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11652971059083939, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0211, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.11875079572200775, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0223, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.11100098490715027, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.022, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.13648055493831635, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0234, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.12870723009109497, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0205, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.10456091910600662, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0232, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.13770641386508942, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0212, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.1526155173778534, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0244, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12664297223091125, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0241, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.12000498175621033, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0239, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.11930232495069504, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.0207, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.11384909600019455, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0217, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.1746685951948166, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.023, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.15868088603019714, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0233, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.15131928026676178, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0194, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.10279352962970734, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0215, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.13810355961322784, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0202, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.13024795055389404, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0217, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1374605894088745, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0215, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.14356477558612823, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0206, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.11046172678470612, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0236, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.10146678239107132, + "learning_rate": 2.601090850395125e-05, + "loss": 0.021, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13247323036193848, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0217, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.13672757148742676, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0201, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.08651759475469589, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.0196, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.14821568131446838, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0223, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11985234171152115, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0208, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11231720447540283, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0219, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.12338588386774063, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0202, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.15079368650913239, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0229, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.1348753422498703, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0209, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1333470344543457, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0207, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.10075798630714417, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0232, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1217522993683815, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0247, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.12041056901216507, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.022, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.10997799783945084, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0226, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.119750015437603, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0239, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.15246307849884033, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0239, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.1431925892829895, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0227, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.1181643158197403, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0227, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.13645906746387482, + "learning_rate": 2.595742267991401e-05, + "loss": 0.0213, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.12661530077457428, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0233, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.1344967484474182, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.0221, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.13433857262134552, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0238, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.10517163574695587, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0237, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.10689525306224823, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.022, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.13307900726795197, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0209, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.14696283638477325, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0261, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.12473653256893158, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0213, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.106630340218544, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0221, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.12092334777116776, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0232, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.11033511906862259, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0193, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11916322261095047, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0212, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.1204444020986557, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0228, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.1373489499092102, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0225, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.13124211132526398, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0208, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.18858365714550018, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0258, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.1558879315853119, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0228, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.1481737494468689, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0242, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.10424906760454178, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.021, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.15494640171527863, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0238, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.1801268309354782, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0238, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.14176428318023682, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0228, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.15850430727005005, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0235, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.14727924764156342, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0232, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.14907293021678925, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0242, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.11011895537376404, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0231, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.13541805744171143, + "learning_rate": 2.588090010443142e-05, + "loss": 0.0224, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.13078846037387848, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0228, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.14508767426013947, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0207, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.12445177882909775, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0227, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.14860141277313232, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0244, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.13311992585659027, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0207, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.1157015860080719, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0209, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.13310302793979645, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.0234, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.14026974141597748, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0246, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.14788632094860077, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0213, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.1551811844110489, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0241, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.10706475377082825, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.021, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.12536053359508514, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0228, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.1260751187801361, + "learning_rate": 2.584384079351628e-05, + "loss": 0.022, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.1082935482263565, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0207, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.13922622799873352, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0251, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1240227147936821, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0223, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.14512665569782257, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0218, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12859801948070526, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.0208, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.13418391346931458, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.021, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.10546918958425522, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.022, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.10668392479419708, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10407276451587677, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.02, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.13980497419834137, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0235, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.16233955323696136, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0224, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.16743852198123932, + "learning_rate": 2.580950866038058e-05, + "loss": 0.022, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.09152300655841827, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0206, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.14319349825382233, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0205, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.1045396700501442, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0219, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.10865644365549088, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0201, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.12166012078523636, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0217, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.12505589425563812, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0224, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.1353408247232437, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0242, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.1379590630531311, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0221, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12337373197078705, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0235, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.12092723697423935, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0233, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.09944337606430054, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0215, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.15161453187465668, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0193, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.0963459238409996, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0225, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.15534570813179016, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0247, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.142191082239151, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0228, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.1454230695962906, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0226, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.12661893665790558, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0211, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.11373333632946014, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0213, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12031935155391693, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.0208, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.13401862978935242, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0224, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.11285468935966492, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0231, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.12453637272119522, + "learning_rate": 2.574625966768285e-05, + "loss": 0.0211, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.13041555881500244, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1391122043132782, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0224, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.1345762014389038, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0244, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.12814867496490479, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0219, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.11058632284402847, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0191, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.15668845176696777, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.021, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.1509203314781189, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.0254, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.13209059834480286, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0201, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.1357070952653885, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0226, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.0896834284067154, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0207, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.1570383608341217, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0234, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.16916486620903015, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0254, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.1385425180196762, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.0218, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.11236828565597534, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0214, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11224918812513351, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0233, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.1305687576532364, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0219, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.15051157772541046, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0226, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.13418416678905487, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0244, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.14324583113193512, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0235, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.10810746997594833, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0197, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.14513953030109406, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0215, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1187531128525734, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0193, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11865831911563873, + "learning_rate": 2.567971342427169e-05, + "loss": 0.021, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.12040061503648758, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0229, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.10597001016139984, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0209, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.1219123974442482, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0242, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.11732123047113419, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.0203, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.15642790496349335, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0226, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.1893257051706314, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0247, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11829031258821487, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0239, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.14971555769443512, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0253, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.12424097210168839, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0251, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1071329414844513, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0271, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.11146610230207443, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0214, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.10925749689340591, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.0212, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.14797845482826233, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0221, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.10348086059093475, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0218, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.14607182145118713, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0217, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.11097954958677292, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0259, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.11300431936979294, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0197, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.11342521011829376, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0202, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.12889035046100616, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0217, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.10449834913015366, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0216, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1293669193983078, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0227, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.10667148977518082, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.021, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.0839434340596199, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0236, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.12906265258789062, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0223, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.1100582480430603, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0243, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.10644520074129105, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.0215, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.11693669110536575, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.022, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.13665053248405457, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0227, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12752027809619904, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0241, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.11534811556339264, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0234, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.1467696875333786, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0245, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10596702992916107, + "learning_rate": 2.558641420481684e-05, + "loss": 0.02, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.14685864746570587, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.022, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.13554386794567108, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.026, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.12572740018367767, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0232, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.1583555042743683, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0209, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.12913484871387482, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0226, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.11418721824884415, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0225, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.12722279131412506, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.0211, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.1382816582918167, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.0243, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.11612103134393692, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0212, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.1537163406610489, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.021, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.10462892800569534, + "learning_rate": 2.555415209239832e-05, + "loss": 0.0214, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.12759479880332947, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0232, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.1290607452392578, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0222, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12458004057407379, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0213, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.1113182008266449, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0223, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.15024979412555695, + "learning_rate": 2.553945541728904e-05, + "loss": 0.022, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.12657828629016876, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.025, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.13719463348388672, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.024, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.13465221226215363, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0254, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.08891251683235168, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0215, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.08946283161640167, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0224, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.13241387903690338, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0229, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.13068708777427673, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0215, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.12115749716758728, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.025, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.12863430380821228, + "learning_rate": 2.551295099554365e-05, + "loss": 0.0202, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.1332341581583023, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0204, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.102240651845932, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0199, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.16963273286819458, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0226, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12840960919857025, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0225, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.09958021342754364, + "learning_rate": 2.549819836273345e-05, + "loss": 0.02, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.10737278312444687, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0226, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.11473958194255829, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0219, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.12188851088285446, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0216, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.1229034960269928, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0227, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.1316317766904831, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1230575442314148, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0218, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13473588228225708, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0221, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1293632537126541, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0223, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.11663107573986053, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0198, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.147003173828125, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0216, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.10070289671421051, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0204, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12855689227581024, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0222, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.10310211777687073, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0227, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.25780248641967773, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0204, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.11621664464473724, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0196, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.15240156650543213, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0232, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.1359490305185318, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0191, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.13072653114795685, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.0242, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.1296282559633255, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0214, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.12183726578950882, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0215, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.10602327436208725, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0222, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.12424091249704361, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0215, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.14287705719470978, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0217, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.10017845034599304, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0214, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.1265304833650589, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0214, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.15279516577720642, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0194, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.11116814613342285, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0197, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.14380785822868347, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0187, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.12522602081298828, + "learning_rate": 2.541224092500406e-05, + "loss": 0.023, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.12386767566204071, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.0197, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11278095096349716, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0205, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12119927257299423, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0206, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.14128951728343964, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0197, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.12576138973236084, + "learning_rate": 2.539735331078111e-05, + "loss": 0.024, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.13276934623718262, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0217, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.15223100781440735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0222, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.13918957114219666, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0211, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.10129881650209427, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0192, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.15048135817050934, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0191, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.11579657346010208, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.019, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.1882474422454834, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0241, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.12608706951141357, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.021, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.14131402969360352, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0234, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.11546603590250015, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0208, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.0989922359585762, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0219, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.13154740631580353, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0234, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.12450987845659256, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0219, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.15696841478347778, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0228, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12059786915779114, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0236, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.1058221235871315, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0234, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.11429761350154877, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0232, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.11841994524002075, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0193, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.12387857586145401, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0239, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.11453676968812943, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0224, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.15571972727775574, + "learning_rate": 2.533461009735874e-05, + "loss": 0.022, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.10241609811782837, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.022, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.11682361364364624, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0211, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.13623486459255219, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0215, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.12484022229909897, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0212, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.09105533361434937, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0223, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.1024160087108612, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0174, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.14108790457248688, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0215, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.1268249899148941, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0211, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12255556881427765, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0192, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.15904900431632996, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0259, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.1356613039970398, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.023, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.12410393357276917, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0223, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1291753053665161, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0221, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.109084852039814, + "learning_rate": 2.529258887162501e-05, + "loss": 0.0243, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.16595998406410217, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0201, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.14062269032001495, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0232, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.13198593258857727, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0214, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.10668856650590897, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0229, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.11266572028398514, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0233, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.12742944061756134, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0221, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.12420892715454102, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0215, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.1265600174665451, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0215, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.11914686858654022, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0222, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.12876294553279877, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0233, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.09800750762224197, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0219, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1273864060640335, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0238, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.13299106061458588, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0226, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.09396592527627945, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.12955208122730255, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0204, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.16461864113807678, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0211, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.11535970121622086, + "learning_rate": 2.524135719213194e-05, + "loss": 0.02, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.11860322207212448, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0212, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10968036204576492, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0207, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.12494095414876938, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.02, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.10398012399673462, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0205, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.12939494848251343, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0216, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.10996326059103012, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0199, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.13704033195972443, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0208, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.11454308778047562, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.0212, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.14186057448387146, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0226, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.12624409794807434, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0199, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.14392296969890594, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.15774424374103546, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0213, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.1352480947971344, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0192, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.10594955086708069, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0231, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.11280091851949692, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0202, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11683343350887299, + "learning_rate": 2.519293379145164e-05, + "loss": 0.022, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.11721377819776535, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0223, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.18298691511154175, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.0225, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12927964329719543, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0202, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.1413726508617401, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.021, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.1330733746290207, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0228, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12341104447841644, + "learning_rate": 2.517472387591168e-05, + "loss": 0.021, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.12199880182743073, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.0202, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.12259527295827866, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.0227, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.11682213097810745, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0246, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.12783272564411163, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0204, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.09663784503936768, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0228, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.13686567544937134, + "learning_rate": 2.515648614306516e-05, + "loss": 0.02, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.12164216488599777, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0234, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.1371183544397354, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0217, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.12031334638595581, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0203, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.1183433011174202, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.1265946477651596, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0224, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.12686678767204285, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0202, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.13161398470401764, + "learning_rate": 2.513517369661726e-05, + "loss": 0.0216, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.13780610263347626, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0205, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.14213374257087708, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0237, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.12532930076122284, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0215, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1276065707206726, + "learning_rate": 2.512297821731217e-05, + "loss": 0.02, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.08893107622861862, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12819649279117584, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0202, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.09707193821668625, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0202, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.511077043763402e-05, + "loss": 0.0208, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.10047636926174164, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0219, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.15637417137622833, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0218, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.15387912094593048, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0213, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.14876650273799896, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0222, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.13631325960159302, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0214, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.12577500939369202, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0232, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15896311402320862, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0237, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.16459719836711884, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0223, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11081109195947647, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0223, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.12327340245246887, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0219, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14812195301055908, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0228, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.11862681806087494, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0252, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.10684223473072052, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0227, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.1279568374156952, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0216, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.13341103494167328, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0213, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.11913862824440002, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0214, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.13493330776691437, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.0214, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.1272144317626953, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0209, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.11348695307970047, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0217, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.16526950895786285, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0198, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.14953579008579254, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0231, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.12980960309505463, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0226, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.16652929782867432, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0244, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1722235530614853, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.026, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.23629626631736755, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0221, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.12417787313461304, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.023, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.11768155544996262, + "learning_rate": 2.502804732645032e-05, + "loss": 0.0186, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.13731959462165833, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.023, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.18034201860427856, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.0201, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11240512877702713, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0217, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.10463474690914154, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0201, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.1273874044418335, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0194, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.12314468622207642, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0237, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.13170552253723145, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.021, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.12725704908370972, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0198, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.11761748790740967, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0198, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13019445538520813, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0237, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.09089519083499908, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.0202, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.13669925928115845, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0236, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.0891583263874054, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0189, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.12015428394079208, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.0214, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.12889859080314636, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0214, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.1092396154999733, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.022, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.1011805608868599, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0201, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.11652060598134995, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.019, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.12280789762735367, + "learning_rate": 2.496950169240971e-05, + "loss": 0.0229, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.1253744661808014, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0245, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.15854160487651825, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0257, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.11671199649572372, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0204, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.15195102989673615, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.11878954619169235, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0207, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.1035882979631424, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0195, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12105780094861984, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0191, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.13278989493846893, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0207, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.16576215624809265, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.0228, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.13104894757270813, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0197, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.14476071298122406, + "learning_rate": 2.49354818242132e-05, + "loss": 0.023, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.14116041362285614, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0204, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.133507639169693, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0213, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11147522926330566, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0192, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.1317354440689087, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0229, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.15885432064533234, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0214, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.0996033102273941, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0213, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.11059842258691788, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0181, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.1103382259607315, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0252, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.12147006392478943, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.1167205348610878, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0217, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.11838674545288086, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0223, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.1745499074459076, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0213, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.09892687201499939, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0213, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.17135944962501526, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0228, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.13930797576904297, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0211, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.10764696449041367, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0204, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11211943626403809, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0207, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1359340399503708, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0215, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.11819283664226532, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0213, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.1166793555021286, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0216, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.14411170780658722, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0206, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.1380920261144638, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0211, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.12901239097118378, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.0224, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.1171313151717186, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.0207, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.15169396996498108, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0225, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.2066943198442459, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0182, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1179543286561966, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0226, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.12273375689983368, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.0224, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.09459412097930908, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0177, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.10208278894424438, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0207, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.11947600543498993, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.0246, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10635806620121002, + "learning_rate": 2.483599695058918e-05, + "loss": 0.021, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.15165211260318756, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0242, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.10573819279670715, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0221, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.10153646767139435, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0212, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.12596352398395538, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.021, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.11259406059980392, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0214, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.09709548950195312, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0186, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1462256908416748, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0228, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.15480898320674896, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0245, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.10707002878189087, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0197, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.10963082313537598, + "learning_rate": 2.480475069778109e-05, + "loss": 0.019, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.13645771145820618, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0229, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.08208916336297989, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0194, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.11223483830690384, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0212, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.12233787775039673, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0228, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.12175288051366806, + "learning_rate": 2.478909963381241e-05, + "loss": 0.0204, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.14267505705356598, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0203, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.12649142742156982, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0217, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.11254285275936127, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0214, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.1454164832830429, + "learning_rate": 2.477656539969673e-05, + "loss": 0.0221, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.17177407443523407, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0212, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.1089746281504631, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0222, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.1525813490152359, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0236, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.09651731699705124, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0214, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.12710145115852356, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0233, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.15999987721443176, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0215, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1600969433784485, + "learning_rate": 2.475460191424065e-05, + "loss": 0.0201, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.11045356094837189, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0208, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.14192573726177216, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.0216, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.11450321227312088, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0202, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.13454747200012207, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.021, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.11518846452236176, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0194, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.10673099011182785, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0214, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.12131883949041367, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0203, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.18098120391368866, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0182, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.09981405735015869, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0214, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.09194957464933395, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0209, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.11133445054292679, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0227, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.14570456743240356, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0181, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.14878204464912415, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0213, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.1475936472415924, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0229, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.14091694355010986, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0229, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.10411489754915237, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.021, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.10015816241502762, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0211, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09339930862188339, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0206, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.13016946613788605, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0223, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.10682770609855652, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0258, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.13658446073532104, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0232, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.12742926180362701, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.13928014039993286, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0208, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.10824346542358398, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0199, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.11346063017845154, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.0225, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12237603217363358, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0226, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.12201637774705887, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0209, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.09991955012083054, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0239, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.15387621521949768, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.025, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.1588744819164276, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0192, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.1347019523382187, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0196, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.12746664881706238, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0212, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11522521078586578, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0197, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.09697534143924713, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0219, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12183485180139542, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0195, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.13119301199913025, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0195, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.10748504102230072, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0192, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.1617487072944641, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0224, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.11518653482198715, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0207, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.17646238207817078, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0233, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.13729365170001984, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.09777814894914627, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0207, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.117672860622406, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0254, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.09014946967363358, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0214, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.1273731142282486, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0223, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10932891070842743, + "learning_rate": 2.460937119575347e-05, + "loss": 0.0224, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.1672651767730713, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0205, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.0930962786078453, + "learning_rate": 2.46030216185615e-05, + "loss": 0.02, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.09708375483751297, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0192, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.13240690529346466, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0214, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.10310725122690201, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0204, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.13352994620800018, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0206, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.15277482569217682, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0222, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.14223086833953857, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0229, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12845414876937866, + "learning_rate": 2.458077513854331e-05, + "loss": 0.0221, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13195961713790894, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0232, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.1461530476808548, + "learning_rate": 2.457441244940383e-05, + "loss": 0.022, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13729079067707062, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0196, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.15615087747573853, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0213, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.12907880544662476, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0198, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.11320316791534424, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0225, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.14982816576957703, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0225, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.12824060022830963, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0196, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.11940523982048035, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0204, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946357995271683, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.0218, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.13296495378017426, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0229, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.1383914053440094, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0207, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.11321834474802017, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0202, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.1276783049106598, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0193, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1406320035457611, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0191, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.12060298770666122, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0195, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.08627335727214813, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0191, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1429424285888672, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0197, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.15133187174797058, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0212, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.11892743408679962, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0231, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.11871744692325592, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0211, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.105750173330307, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.0234, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10969540476799011, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.022, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.09607745707035065, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0173, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.11796857416629791, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0193, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.114372618496418, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.02, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.1164821982383728, + "learning_rate": 2.449463406055895e-05, + "loss": 0.023, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.13561253249645233, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.018, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.14956215023994446, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0245, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.12785622477531433, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0222, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.16085495054721832, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.021, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10792174935340881, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0209, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.10508404672145844, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0217, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.11471311002969742, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0202, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.14265185594558716, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0224, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.13569262623786926, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0196, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15635624527931213, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0205, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12042741477489471, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0179, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.09488346427679062, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0198, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.20503900945186615, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0199, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.13667535781860352, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13254515826702118, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0204, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.13879507780075073, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0233, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.11448678374290466, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0237, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1255042403936386, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0204, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11896912008523941, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0207, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.12624827027320862, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0194, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.09312308579683304, + "learning_rate": 2.442727178914008e-05, + "loss": 0.022, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12632645666599274, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0189, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.11468527466058731, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0182, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.09245705604553223, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0188, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.12113099545240402, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0191, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.13049040734767914, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0221, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.16979938745498657, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0211, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.11175577342510223, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0172, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.1191241443157196, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0206, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.16317729651927948, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.0213, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.10087717324495316, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0194, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.12195733189582825, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0253, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11880146712064743, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0222, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.14213789999485016, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0198, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.15624478459358215, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.0205, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09695806354284286, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0219, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11260177940130234, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0206, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.15394112467765808, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0219, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.11663617193698883, + "learning_rate": 2.436928131964383e-05, + "loss": 0.0214, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.12596042454242706, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0225, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.10310690104961395, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0203, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.12332364916801453, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0203, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.16257967054843903, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0234, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.13102053105831146, + "learning_rate": 2.435313188161029e-05, + "loss": 0.022, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.14736537635326385, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0199, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.14611469209194183, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0238, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.14594200253486633, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0212, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.16921328008174896, + "learning_rate": 2.434019954415699e-05, + "loss": 0.0236, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.1318638175725937, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0201, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.10869060456752777, + "learning_rate": 2.433372911848899e-05, + "loss": 0.019, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.1446760594844818, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.021, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.1298162341117859, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10307328402996063, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0197, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.11932921409606934, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0209, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.10986137390136719, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0227, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.12099824845790863, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0182, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.13432320952415466, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.0213, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.14018790423870087, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0199, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.13015322387218475, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0191, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.10281360149383545, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0215, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.09500012546777725, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0216, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.15131708979606628, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0225, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.08750247210264206, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0221, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.10143480449914932, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0214, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.12935201823711395, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0202, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.10703578591346741, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0207, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08649012446403503, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0184, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.11186458170413971, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.021, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.11181459575891495, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0211, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.11756999045610428, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0216, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.12061341106891632, + "learning_rate": 2.426561884159603e-05, + "loss": 0.019, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.13861820101737976, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0214, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.11545911431312561, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0195, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.14892302453517914, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0226, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.10354511439800262, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0197, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.15290679037570953, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0208, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.12185964733362198, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0205, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.10486926138401031, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0206, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.10959216207265854, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0198, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.11638635396957397, + "learning_rate": 2.423633359536448e-05, + "loss": 0.0227, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.13661672174930573, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0206, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.13071168959140778, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0189, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.16377446055412292, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0203, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.22221329808235168, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0185, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.11251460015773773, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0191, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.10873523354530334, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0228, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.13705629110336304, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0203, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.13980528712272644, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.10355913639068604, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0199, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.1282302737236023, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.0215, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.1221374049782753, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0192, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.18542620539665222, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0199, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1250954121351242, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0189, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12665164470672607, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0217, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.09464617818593979, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.0197, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.09657798707485199, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0194, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.11533467471599579, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0189, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.13647402822971344, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0194, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.13276426494121552, + "learning_rate": 2.41743228236291e-05, + "loss": 0.022, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.16899874806404114, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0209, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.14313359558582306, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0211, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.14507170021533966, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0218, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.13025349378585815, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0209, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.1666184365749359, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0217, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14891882240772247, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0205, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11392978578805923, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0223, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.12964752316474915, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.019, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11125730723142624, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0211, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.08378217369318008, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0203, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.14530996978282928, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0206, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.10187243670225143, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0188, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.09652446955442429, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0192, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.10552513599395752, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0212, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11044801771640778, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0215, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.12248426675796509, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0185, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.11164992302656174, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0198, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.11919678002595901, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0187, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1675145924091339, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0262, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.15451745688915253, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0226, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.0997307077050209, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0196, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.1272345781326294, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0214, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.1286081224679947, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0214, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12226564437150955, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0211, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.11795010417699814, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0216, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.12963736057281494, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0215, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.10415369272232056, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0215, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.11642655730247498, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0194, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.19006073474884033, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0216, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.16677096486091614, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.0235, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.15928123891353607, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0195, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.10270075500011444, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0197, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.12253808975219727, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0216, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14145877957344055, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0211, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1387409120798111, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0185, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.11433693021535873, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0204, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.14247572422027588, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0187, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.12203793972730637, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.018, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.15705470740795135, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0212, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.10648002475500107, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0182, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.11573450267314911, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0211, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.11960671842098236, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0245, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.14416365325450897, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0178, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.11234421283006668, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0205, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.1375260204076767, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.025, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.08686859905719757, + "learning_rate": 2.402315322351119e-05, + "loss": 0.018, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11352542787790298, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0198, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.13374295830726624, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0216, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.09358088672161102, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0202, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.13767337799072266, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0229, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.1405765861272812, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.02, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11317892372608185, + "learning_rate": 2.400332803040664e-05, + "loss": 0.02, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.12356667965650558, + "learning_rate": 2.400002143698994e-05, + "loss": 0.0225, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.12570610642433167, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0232, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.1461946964263916, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0201, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.10893318802118301, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0211, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.10969080775976181, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0235, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.17320629954338074, + "learning_rate": 2.398347822254704e-05, + "loss": 0.0221, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14296430349349976, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0206, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12349261343479156, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0182, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.11617975682020187, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0209, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.1574973613023758, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0222, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.157145157456398, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.0199, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.14698748290538788, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0203, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.12345635890960693, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0212, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.1370248794555664, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0194, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.1332298070192337, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0219, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.12525111436843872, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.0219, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.14905311167240143, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0211, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12063094973564148, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.0262, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.10361988842487335, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0195, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.09919667989015579, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0201, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.11141914129257202, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.0223, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.17785966396331787, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0201, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.126350998878479, + "learning_rate": 2.392710390356112e-05, + "loss": 0.0201, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.12107345461845398, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0213, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.11466097086668015, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0202, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.15397872030735016, + "learning_rate": 2.391713512040237e-05, + "loss": 0.022, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.14412778615951538, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0223, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.10495682060718536, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0194, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.13021183013916016, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0199, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.13993357121944427, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0182, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.128571018576622, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.0181, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.11090676486492157, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0204, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.10988310724496841, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0203, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.11705116927623749, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0214, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.1277252584695816, + "learning_rate": 2.388719222953459e-05, + "loss": 0.0225, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1561414748430252, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0221, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.15278629958629608, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.02, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14865094423294067, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0213, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.1127776950597763, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0211, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.15085943043231964, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0212, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14555132389068604, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0216, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.14708220958709717, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0209, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.1093590185046196, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0199, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.1303861141204834, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0242, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.11462346464395523, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0197, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.12198365479707718, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0206, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.11193065345287323, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0201, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.13254034519195557, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0181, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.13355755805969238, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0204, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12551282346248627, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0212, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1441735178232193, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0217, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.11511780321598053, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0218, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.10427968204021454, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0222, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.11364301294088364, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0166, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.1401003897190094, + "learning_rate": 2.38204570134882e-05, + "loss": 0.0215, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.11645112931728363, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0198, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1340353935956955, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0212, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.0982050895690918, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0208, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.16713382303714752, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0185, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.13803045451641083, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.021, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.15552546083927155, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0233, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.10836438834667206, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0211, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.1831241250038147, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0198, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.12369874119758606, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0193, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.10911241918802261, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0221, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.11441849172115326, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0236, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.11105310916900635, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0206, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.13617953658103943, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0189, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.11214766651391983, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0192, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.12536488473415375, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0206, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.10356712341308594, + "learning_rate": 2.376687582517292e-05, + "loss": 0.019, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.13332195580005646, + "learning_rate": 2.376352133090187e-05, + "loss": 0.0216, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.13996818661689758, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0216, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.14305290579795837, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0204, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.10494769364595413, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0202, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.1461218297481537, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.0197, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.12469783425331116, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0207, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.13164709508419037, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.0207, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.1553804576396942, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0228, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1445176750421524, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0217, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12673451006412506, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0164, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.17502479255199432, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0208, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.11328406631946564, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0196, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12234057486057281, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0193, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.1333971917629242, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0214, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.11227905005216599, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0183, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.1287100911140442, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0198, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.10215240716934204, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0194, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.09786848723888397, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0202, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.11103200167417526, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0219, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.1491294801235199, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0202, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.1364760547876358, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.0198, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.11583423614501953, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0206, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.13896186649799347, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0183, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.0970211774110794, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.019, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.10326618701219559, + "learning_rate": 2.368281439559657e-05, + "loss": 0.022, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.11581890285015106, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0234, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10186298936605453, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0205, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11655393242835999, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0193, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.11990638077259064, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0217, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.10573728382587433, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0177, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.09257248789072037, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.0216, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.11310117691755295, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0188, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.12632012367248535, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0217, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1535339206457138, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0209, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.13133156299591064, + "learning_rate": 2.364907420469738e-05, + "loss": 0.0207, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11225332319736481, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0217, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.10744500905275345, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0212, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.10509403049945831, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0208, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.16215650737285614, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0199, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1208849623799324, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0196, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.14882729947566986, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0229, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.18775896728038788, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0206, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.14199234545230865, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0201, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.1510499119758606, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.021, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.16389605402946472, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0197, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.13032132387161255, + "learning_rate": 2.361188413960337e-05, + "loss": 0.0211, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.13895465433597565, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0222, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.10914628207683563, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0206, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.12215261161327362, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0186, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12954379618167877, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.0193, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.12779726088047028, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.02, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.11825352907180786, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0214, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11001764982938766, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0211, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.10464788973331451, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.02, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.0892137810587883, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0198, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.09811265021562576, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0191, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.14199955761432648, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0196, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.13846102356910706, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0216, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.12102426588535309, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0214, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.12360251694917679, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0199, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.12587665021419525, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.15562282502651215, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.0214, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.11366713047027588, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0192, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.15177540481090546, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0204, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.08908293396234512, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0186, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.1349143236875534, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0197, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.09388674795627594, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0205, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.11973906308412552, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0198, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.1290522664785385, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.02, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.12087979167699814, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0207, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.10237318277359009, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0189, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.10426577180624008, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0204, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13028934597969055, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0197, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.10696206241846085, + "learning_rate": 2.351686222335574e-05, + "loss": 0.0206, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.11024264991283417, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0189, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.11294405907392502, + "learning_rate": 2.35100554725415e-05, + "loss": 0.0228, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.1188240721821785, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0204, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.17485342919826508, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0202, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12582673132419586, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0209, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.18180538713932037, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0202, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.1459868997335434, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0189, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.09818759560585022, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0195, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.1473611295223236, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0253, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.09137327969074249, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0209, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.14026956260204315, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0202, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.1511334776878357, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0182, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.15016451478004456, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0204, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.14407901465892792, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.0205, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.1332983374595642, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0192, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.14312349259853363, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.0195, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.11917808651924133, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0194, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.17896629869937897, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0203, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.12084568291902542, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0199, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.14310620725154877, + "learning_rate": 2.344867863274048e-05, + "loss": 0.0201, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1168353408575058, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0225, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.14718633890151978, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0204, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.14102597534656525, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0197, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.09543738514184952, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0218, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1074792668223381, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0198, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.11555791646242142, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0196, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.12645067274570465, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0191, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.11634501069784164, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.02, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.09201087802648544, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0204, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.1155405044555664, + "learning_rate": 2.341449046471072e-05, + "loss": 0.0215, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.09871169179677963, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0216, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.11332481354475021, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0179, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.12474322319030762, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.0177, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.11339092999696732, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0221, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.1200975552201271, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.0209, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.14456698298454285, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0211, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.1371302306652069, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0165, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.11727641522884369, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0232, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.11336934566497803, + "learning_rate": 2.338366646946928e-05, + "loss": 0.0202, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.10325363278388977, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0209, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.11268003284931183, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0187, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10260385274887085, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.0195, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.10920321941375732, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0225, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.09120934456586838, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0194, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11567433178424835, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0197, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.10144805908203125, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0206, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1246950775384903, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0215, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.100107841193676, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0197, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.15822021663188934, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0202, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13518433272838593, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.02, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12268074601888657, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0196, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.12333078682422638, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0184, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.13588106632232666, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0218, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.12120439857244492, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0197, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.16428445279598236, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0198, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.11615289747714996, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.0182, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.1432264745235443, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0207, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.11717039346694946, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0198, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.10115417838096619, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0193, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.11894490569829941, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0198, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.11940047889947891, + "learning_rate": 2.330810220141532e-05, + "loss": 0.021, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.10992413759231567, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0195, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.10544339567422867, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0182, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.11725642532110214, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.0198, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.09130900353193283, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0209, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.09033828228712082, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0177, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.11721840500831604, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0207, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.12251118570566177, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0215, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.1496249884366989, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0219, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.14808881282806396, + "learning_rate": 2.327710137160376e-05, + "loss": 0.0212, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.145681232213974, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.019, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.12777844071388245, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0173, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.1110587939620018, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0198, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.11348012834787369, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0207, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.11685867607593536, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.02, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.17935989797115326, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0211, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.12398742884397507, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0191, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.10836582630872726, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0196, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.12038073688745499, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0207, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.12211567908525467, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0206, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.1702142357826233, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0227, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12037540227174759, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0191, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.15964926779270172, + "learning_rate": 2.323223255665661e-05, + "loss": 0.018, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.12493535876274109, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0212, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.1359032839536667, + "learning_rate": 2.32253202698687e-05, + "loss": 0.0219, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.18753379583358765, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0228, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.1336824893951416, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0201, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.09901099652051926, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0221, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.12925978004932404, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0206, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.1278320550918579, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.0178, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.12977559864521027, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0176, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10098735243082047, + "learning_rate": 2.320110760314603e-05, + "loss": 0.023, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.12594632804393768, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.0209, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.1140967383980751, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0181, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.1287144273519516, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0195, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.12329071760177612, + "learning_rate": 2.318725808577064e-05, + "loss": 0.0198, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.1339070051908493, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10581240057945251, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0265, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.13716311752796173, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0213, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.14610108733177185, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0195, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.128855898976326, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0185, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.0950121283531189, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0223, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.14444753527641296, + "learning_rate": 2.316299750241968e-05, + "loss": 0.017, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.12585613131523132, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0177, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.1290416419506073, + "learning_rate": 2.315606032389525e-05, + "loss": 0.02, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.10740494728088379, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0191, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11578185111284256, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0211, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.11840540170669556, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0209, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.09531640261411667, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0202, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.10436122864484787, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0188, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.14102807641029358, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0205, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14070506393909454, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0186, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.09883972257375717, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0196, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11524099856615067, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0187, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.11240902543067932, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0176, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.08062270283699036, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.019, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.13965697586536407, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0202, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.11376950889825821, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0218, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.13540907204151154, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0188, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.10888075083494186, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0216, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.10280206799507141, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0178, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.11512117087841034, + "learning_rate": 2.309699448641602e-05, + "loss": 0.02, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.09106973558664322, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0182, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10531709343194962, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0204, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.159554585814476, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0173, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.12161346524953842, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0196, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.08224594593048096, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0175, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.12767191231250763, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0234, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.12864384055137634, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0224, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.10232329368591309, + "learning_rate": 2.3069137227667e-05, + "loss": 0.0208, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.09609836339950562, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0183, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.1301981806755066, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0202, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.1258714497089386, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0217, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.12543867528438568, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13326771557331085, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.022, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.11770334839820862, + "learning_rate": 2.304821853869955e-05, + "loss": 0.019, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.11729256063699722, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0187, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.11887779086828232, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0224, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.10921293497085571, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0192, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.1346455216407776, + "learning_rate": 2.303426051834282e-05, + "loss": 0.022, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.1177918091416359, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0209, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.12912803888320923, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0191, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.12571996450424194, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0211, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.13959148526191711, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0198, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.1404334306716919, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0211, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.1111592948436737, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0213, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.10322082787752151, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0203, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.11235317587852478, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0187, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.13599546253681183, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0205, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11067865788936615, + "learning_rate": 2.299932279415038e-05, + "loss": 0.02, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.14130094647407532, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0204, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.14370687305927277, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0199, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.13825304806232452, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0197, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.11885961890220642, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.02, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.12376029789447784, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0199, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.10103051364421844, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0181, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.10931621491909027, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0198, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.11835317313671112, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0202, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.08036374300718307, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0189, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.08135218918323517, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0196, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.12611854076385498, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0211, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.09217566251754761, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0181, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.10258878767490387, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0194, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.1165921688079834, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0196, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.09290144592523575, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.0205, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.11575523763895035, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0204, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13866478204727173, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0178, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.13036414980888367, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0191, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11062757670879364, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0193, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.09973529726266861, + "learning_rate": 2.292926536293724e-05, + "loss": 0.02, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.13707594573497772, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0186, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11635670065879822, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0185, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.13926011323928833, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0192, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.13966591656208038, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0194, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.09878037869930267, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0191, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.11935862898826599, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0193, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.15291833877563477, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0189, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.12976837158203125, + "learning_rate": 2.290117482755701e-05, + "loss": 0.017, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.13367557525634766, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0178, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12121888250112534, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0191, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.13135644793510437, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0226, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.1142699271440506, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.017, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.09670589119195938, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0199, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.12751951813697815, + "learning_rate": 2.288008171679078e-05, + "loss": 0.018, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.14463315904140472, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0184, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.11006469279527664, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0198, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.15361396968364716, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0186, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.20782725512981415, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0208, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.15324300527572632, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.018, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.17235179245471954, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0218, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.12993113696575165, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0198, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.1496446281671524, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0203, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.16271376609802246, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.0215, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11106697469949722, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0206, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.10508760809898376, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0218, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.11433881521224976, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0203, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.12877102196216583, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0203, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.13301214575767517, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0225, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.15351711213588715, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0213, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.1114102303981781, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0212, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.09978143125772476, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0198, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10524465143680573, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0183, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.09780002385377884, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0185, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.1361900269985199, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0218, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.1329333484172821, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0191, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.10501913726329803, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.019, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.1329827904701233, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0198, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.10849577933549881, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0213, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.14308017492294312, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.0199, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.12313494831323624, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0193, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.10307908058166504, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.0215, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10591195523738861, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0178, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.1411430686712265, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0237, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13116225600242615, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0203, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.12032114714384079, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0201, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.12519164383411407, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0177, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.10340812057256699, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0194, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.10876207798719406, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0207, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11245796829462051, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0207, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11444602906703949, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0171, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12232744693756104, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0196, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.10049490630626678, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0179, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.09973384439945221, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.022, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.11037951707839966, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0197, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.14473563432693481, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0207, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.1138916090130806, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0198, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.12471994757652283, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0182, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.12418543547391891, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0195, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.12200118601322174, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.0217, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.13389603793621063, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0194, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10176926106214523, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.0189, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.10195910185575485, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0202, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.13289214670658112, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.018, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.10307228565216064, + "learning_rate": 2.270347347405331e-05, + "loss": 0.0183, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.149918794631958, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0172, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.1182866096496582, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0173, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.1035924181342125, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0194, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.108797088265419, + "learning_rate": 2.268928118647676e-05, + "loss": 0.0174, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.11457139253616333, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0208, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.10256518423557281, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0188, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13053470849990845, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0209, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.11632310599088669, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0187, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.12769649922847748, + "learning_rate": 2.267152769015067e-05, + "loss": 0.0206, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.13292407989501953, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0191, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.09822595119476318, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0189, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.13729946315288544, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0229, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.15705017745494843, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0203, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.12355367839336395, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0201, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.1278308629989624, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0205, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.09874208271503448, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0218, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.1316995620727539, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0212, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.10361800342798233, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0209, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.09672107547521591, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0205, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.12609392404556274, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0205, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.1360992044210434, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0188, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.11627763509750366, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0206, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.11733578145503998, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.02, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.1705910563468933, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0174, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.11704754829406738, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0199, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.13247141242027283, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0224, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.10916013270616531, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0202, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.11092681437730789, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.0186, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.12811405956745148, + "learning_rate": 2.260036839187291e-05, + "loss": 0.02, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.11893392354249954, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0194, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.16553092002868652, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0189, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.12431131303310394, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.02, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10850611329078674, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0222, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.12647055089473724, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0193, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.10432784259319305, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0196, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.1159503236413002, + "learning_rate": 2.257540798155951e-05, + "loss": 0.0192, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.10592516511678696, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0219, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.1318269520998001, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0195, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.11932560801506042, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0194, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.16499531269073486, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.02, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.1334158480167389, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.0217, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.14781776070594788, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0206, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.14816460013389587, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0206, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.11319883167743683, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0196, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.12413972616195679, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0206, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.1277540624141693, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0211, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.12178509682416916, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0192, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.1210661232471466, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0232, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.11300696432590485, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0182, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.15876832604408264, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0194, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.1387646645307541, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0202, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.12502169609069824, + "learning_rate": 2.251824990920905e-05, + "loss": 0.02, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.12365051358938217, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0197, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.12610606849193573, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0194, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.13096576929092407, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0208, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.13002553582191467, + "learning_rate": 2.250393750922939e-05, + "loss": 0.02, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.10048994421958923, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0193, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.11463778465986252, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0179, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10637608915567398, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0182, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.11959703266620636, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0183, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.11852603405714035, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12075617164373398, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0186, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.09626025706529617, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0195, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12773670256137848, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.019, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13364142179489136, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.019, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.14562146365642548, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0207, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.11752697080373764, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.13970085978507996, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0219, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.12456551194190979, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0203, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.36701101064682007, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0191, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.17254851758480072, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0193, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.1397981345653534, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0205, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.1273665875196457, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0227, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.08652572333812714, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0194, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.118954136967659, + "learning_rate": 2.243582938893974e-05, + "loss": 0.019, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.12105695903301239, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0173, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.10059279948472977, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0189, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.1476186215877533, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0192, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.11198431998491287, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0212, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.14829197525978088, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0216, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.14479608833789825, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0225, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.16737256944179535, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0202, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.12101127207279205, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0198, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.09408389776945114, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.019, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.10649813711643219, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0191, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.1328127533197403, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.02, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1135721281170845, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0181, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.14464536309242249, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0167, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.15146523714065552, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0216, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11212945729494095, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0189, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.11722676455974579, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0185, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10031717270612717, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0181, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.14852827787399292, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.02, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12114712595939636, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0207, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.12325385212898254, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0185, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.12915199995040894, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0188, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.10423776507377625, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0195, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.13398106396198273, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0182, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.1426725685596466, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0214, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.12515439093112946, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0184, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.15888792276382446, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0183, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.11037078499794006, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0199, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.10706029832363129, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.0195, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.11116144061088562, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0202, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.12716875970363617, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0182, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.1446460634469986, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0172, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.09947994351387024, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0205, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.09251125156879425, + "learning_rate": 2.231705291760663e-05, + "loss": 0.018, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.08440345525741577, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0181, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12813445925712585, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0206, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.1017315611243248, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0192, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.13247326016426086, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0212, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.14845387637615204, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0194, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.13824601471424103, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0224, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.1151830404996872, + "learning_rate": 2.229177978101565e-05, + "loss": 0.0176, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.12420724332332611, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0179, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.13085328042507172, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0179, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.12436439841985703, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.11873050034046173, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0197, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.1338828206062317, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0212, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.1106492280960083, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0179, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.16542823612689972, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0202, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.08829429000616074, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0176, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.10792568325996399, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0186, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.1279839128255844, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0221, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.10742732882499695, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0217, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.10220900177955627, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0201, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.1660427749156952, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.019, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.13747647404670715, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0225, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.13351675868034363, + "learning_rate": 2.223753180973339e-05, + "loss": 0.021, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.14151513576507568, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.0195, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.12823058664798737, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0189, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.12779751420021057, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0177, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.2243196815252304, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0183, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.12365520745515823, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0188, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.10981941968202591, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0191, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.11569637805223465, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0195, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.15785852074623108, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0222, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11610189825296402, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0172, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.1124223843216896, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0184, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.14634257555007935, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.0196, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.13306891918182373, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0204, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.08198657631874084, + "learning_rate": 2.219041683306619e-05, + "loss": 0.0177, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.09529033303260803, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0166, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.11170824617147446, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0195, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13825464248657227, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0191, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.11621235311031342, + "learning_rate": 2.217590132288247e-05, + "loss": 0.0202, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11921709030866623, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.019, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12485305964946747, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0205, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.1244436725974083, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0204, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11240405589342117, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0197, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.1087997704744339, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0183, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.11403695493936539, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0172, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.12764202058315277, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0161, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.17418256402015686, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0171, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.1522650271654129, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0197, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.16773882508277893, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0214, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.14681261777877808, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0208, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.14842623472213745, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.021, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.10614728182554245, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0182, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.15053948760032654, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0209, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.1381484866142273, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.0176, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.13161014020442963, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0174, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.09695997089147568, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0202, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.08903522044420242, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0171, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11456362903118134, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0183, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.1260307878255844, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.02, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.11732659488916397, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0197, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.126072496175766, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0196, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.11586430668830872, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.02, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.10504680871963501, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0193, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.09652367234230042, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0186, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10370496660470963, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0176, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.1192321851849556, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0178, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.10527302324771881, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0182, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10162831097841263, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0191, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.16128556430339813, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0203, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.11897512525320053, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0204, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.14880484342575073, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0205, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.11369120329618454, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.0193, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.11618658900260925, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0176, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.10052158683538437, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.0207, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1014244481921196, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0205, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.15118759870529175, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0215, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.09245041012763977, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0176, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.13509149849414825, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0224, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.12747371196746826, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0185, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11445175856351852, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0208, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11949758231639862, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0193, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.12797918915748596, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0182, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0185, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.11956185847520828, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0189, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.11755053699016571, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0184, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.1273607611656189, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0205, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.11855774372816086, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0188, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.09031623601913452, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0183, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12208469212055206, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0199, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.1083436831831932, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0193, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12228160351514816, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0182, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.1007765457034111, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0195, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.11114434152841568, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0173, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.1096438392996788, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0199, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.124867744743824, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0219, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.10032925009727478, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.0194, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.13164062798023224, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0193, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.1376691609621048, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0187, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.11254525184631348, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0186, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.12835443019866943, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.15313562750816345, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0219, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.16989216208457947, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0172, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.11433102935552597, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.019, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09817333519458771, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0222, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.09500890970230103, + "learning_rate": 2.193514973491427e-05, + "loss": 0.0176, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.14314892888069153, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0199, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.11764178425073624, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0169, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.12384455651044846, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0193, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.11438048630952835, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0174, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.13921929895877838, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0182, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.16880372166633606, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0174, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.1554516702890396, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0194, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.15355393290519714, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0212, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.1179167702794075, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0195, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.13388752937316895, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0208, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.10997489839792252, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0181, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.11783649772405624, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0189, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13451549410820007, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0186, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.1339322179555893, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0179, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.09972941130399704, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0214, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.10101358592510223, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0176, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.09771978110074997, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0177, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.14234617352485657, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0184, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.10462982952594757, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0197, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.09772231429815292, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0226, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.12018870562314987, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0164, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.10980863869190216, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0199, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.11285397410392761, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0172, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.1067347601056099, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0183, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.1046559289097786, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.0198, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.11118492484092712, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0179, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.1001342311501503, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0166, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.1053142324090004, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0188, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.16253869235515594, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.018, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.10042452812194824, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0173, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.11815430223941803, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0181, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.07585619390010834, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0195, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.17831926047801971, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0183, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.09953836351633072, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0194, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.12129344046115875, + "learning_rate": 2.18065424687869e-05, + "loss": 0.0188, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.09982063621282578, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0196, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.0906965360045433, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0172, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.1256277859210968, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.019, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.10290617495775223, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0171, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.17007306218147278, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0178, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.0779242068529129, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0192, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.12276264280080795, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0233, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.13211409747600555, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0184, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11407056450843811, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0211, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.13331526517868042, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0193, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.1603303849697113, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0192, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.10634022206068039, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0174, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.1658177524805069, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0207, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.1432579606771469, + "learning_rate": 2.175492159090619e-05, + "loss": 0.02, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.11264102905988693, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0183, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.12988954782485962, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.0199, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.1687532663345337, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0175, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12261682003736496, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0148, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.13096439838409424, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0175, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.13871102035045624, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0198, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.12536168098449707, + "learning_rate": 2.172907339903944e-05, + "loss": 0.018, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.12913931906223297, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0194, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.1255621761083603, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0172, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.12288706004619598, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0206, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.113137386739254, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0201, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.11107201129198074, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.14609375596046448, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.02, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.15227735042572021, + "learning_rate": 2.170320016677318e-05, + "loss": 0.016, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.12586566805839539, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0194, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.13495711982250214, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0187, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.11436781287193298, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.018, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.12271122634410858, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0193, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.16221405565738678, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0195, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.08645932376384735, + "learning_rate": 2.168100325341058e-05, + "loss": 0.0157, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.11750698834657669, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0186, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.1191735714673996, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0178, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.10067173838615417, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0187, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.09761539846658707, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0169, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.13806267082691193, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0217, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.10618548095226288, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0165, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.15135695040225983, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0199, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10563432425260544, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0176, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.12757790088653564, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0183, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.11258719861507416, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.0162, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.15290242433547974, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0192, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.1350955218076706, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0188, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.09494426846504211, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0187, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12816165387630463, + "learning_rate": 2.162913952762737e-05, + "loss": 0.019, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.1237507089972496, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0187, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.12867869436740875, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0202, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11882596462965012, + "learning_rate": 2.161801300808089e-05, + "loss": 0.019, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.10291648656129837, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0187, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.09269602596759796, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.019, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.08865118771791458, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.0188, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10444393008947372, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0182, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.11461203545331955, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.0219, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.13728564977645874, + "learning_rate": 2.159574640654742e-05, + "loss": 0.0223, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.1394069343805313, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0172, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.15394151210784912, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0182, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.10422483086585999, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0175, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.10996434837579727, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0182, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.1026761382818222, + "learning_rate": 2.157717712784381e-05, + "loss": 0.018, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.1373908817768097, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0226, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.10155291855335236, + "learning_rate": 2.156974591793542e-05, + "loss": 0.0179, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.14904215931892395, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0192, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.1131916344165802, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0175, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.09735766798257828, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0182, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.13132493197917938, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0193, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.12934266030788422, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0184, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.1196565330028534, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0186, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.12375814467668533, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.0199, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.0892246812582016, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0178, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.0932748094201088, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0192, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11503782123327255, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0187, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.08250989019870758, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0183, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.101202592253685, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0183, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1226576566696167, + "learning_rate": 2.152139450890464e-05, + "loss": 0.022, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.1167614683508873, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0176, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.12396091967821121, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0186, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.13461633026599884, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.0169, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.10792742669582367, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0189, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11649876832962036, + "learning_rate": 2.150277549265773e-05, + "loss": 0.02, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.1314917951822281, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0157, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.14143578708171844, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0197, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10042720288038254, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0177, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.12019779533147812, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0164, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.14096254110336304, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0223, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.10627058893442154, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0171, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.11092417687177658, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.019, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.12211407721042633, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0199, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11551571637392044, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0197, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.10474767535924911, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.018, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.12364303320646286, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0178, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13243450224399567, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0197, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10594736039638519, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0179, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.11143538355827332, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0196, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.10518378764390945, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0197, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12268701940774918, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0176, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.09408178925514221, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0205, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.12529458105564117, + "learning_rate": 2.143564506261267e-05, + "loss": 0.0184, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.15819092094898224, + "learning_rate": 2.143191093594534e-05, + "loss": 0.021, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.15818199515342712, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0182, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.13374921679496765, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0177, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.10046645253896713, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0197, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.11422070860862732, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0185, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.1138019785284996, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0182, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.09617039561271667, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0178, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.1452113538980484, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0222, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.10409227013587952, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0196, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.11277221143245697, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0175, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.12325838208198547, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0201, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12721778452396393, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0174, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.14016005396842957, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.017, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.11149191111326218, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0168, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.13445520401000977, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0194, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10086208581924438, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0197, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.115168996155262, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0181, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.11483300477266312, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.0183, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.1196938082575798, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0172, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.12219196557998657, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0177, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.11318044364452362, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.0204, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.09219531714916229, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0181, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.09911682456731796, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0192, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.11822745203971863, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0192, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.11101230978965759, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0165, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.09978491812944412, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0203, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.09874484688043594, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0199, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.08682837337255478, + "learning_rate": 2.133090580749213e-05, + "loss": 0.0177, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.14352355897426605, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0189, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.11133426427841187, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0199, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.10667043924331665, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0183, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.14207594096660614, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0218, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.12165568023920059, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0167, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.09311263263225555, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0183, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10423581302165985, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0151, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.12640471756458282, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0214, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.11349574476480484, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0171, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.14359329640865326, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0182, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.10977838933467865, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0181, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.12443531304597855, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0168, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.12441008538007736, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0186, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.14103129506111145, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0197, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10460726171731949, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0178, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.12526848912239075, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0188, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.13363681733608246, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0183, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.13106584548950195, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0181, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.1353096216917038, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0202, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.11889169365167618, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0198, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.12628430128097534, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0178, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.13260240852832794, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0182, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12357809394598007, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0175, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.13913387060165405, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0189, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.12303008884191513, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0187, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.10231908410787582, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.017, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.20065239071846008, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0172, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.11311347782611847, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.019, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.131951242685318, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0182, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.15945149958133698, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0202, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09272884577512741, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0189, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.0937163308262825, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0208, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.10963460803031921, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0197, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.1618727743625641, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0195, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.0919048935174942, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0177, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.12060586363077164, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0209, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.13121892511844635, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0181, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.11795026063919067, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0186, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.13060589134693146, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0226, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.08142203092575073, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0181, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.12634120881557465, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0178, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.1448773890733719, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0179, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1230989471077919, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0188, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.11944684386253357, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0212, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.1445692479610443, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0184, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.0831015482544899, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0171, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.1456967294216156, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0189, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.11207464337348938, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0185, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.16137553751468658, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0182, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.09820980578660965, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0192, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.11823676526546478, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0183, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.12873874604701996, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0181, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.11243417114019394, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0188, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.11120739579200745, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0222, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.10704528540372849, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0178, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.12063440680503845, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0174, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.13275623321533203, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0171, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.10400331765413284, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0201, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.10485411435365677, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0187, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.12517869472503662, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0187, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.11703156679868698, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.0178, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.11982452869415283, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0186, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11858765780925751, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0182, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.10404367744922638, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0173, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.09697552025318146, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0188, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.12046534568071365, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0173, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.0896986722946167, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0188, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.10993069410324097, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0213, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.1200791671872139, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0179, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.14664512872695923, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.0202, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.09285914152860641, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0207, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.1080656424164772, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0188, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.140005961060524, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0188, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.11084222793579102, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.02, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.0913061872124672, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0194, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.09459970891475677, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0182, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.09709502756595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0195, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.11101861298084259, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0184, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.1152782216668129, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0178, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.11495838314294815, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0178, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.14250393211841583, + "learning_rate": 2.10258099183954e-05, + "loss": 0.019, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.09898682683706284, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0173, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.13904057443141937, + "learning_rate": 2.101823807728166e-05, + "loss": 0.018, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.12013959139585495, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0163, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.0925723984837532, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0187, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.1284961700439453, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0158, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.12408998608589172, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0206, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1095876470208168, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0172, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.10166801512241364, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0183, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13473083078861237, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0155, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.10492225736379623, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0188, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.10769215226173401, + "learning_rate": 2.098414220174082e-05, + "loss": 0.021, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.09331739693880081, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0176, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.09247364103794098, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0163, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.20951682329177856, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0203, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.16750548779964447, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0196, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.10933154821395874, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0169, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10767697542905807, + "learning_rate": 2.09613911545763e-05, + "loss": 0.0184, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.11565674096345901, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0208, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.11934554576873779, + "learning_rate": 2.095380384728948e-05, + "loss": 0.018, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.12786556780338287, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0169, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.11752485483884811, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0194, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.11586920917034149, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0179, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.08849998563528061, + "learning_rate": 2.0938623809196e-05, + "loss": 0.018, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.12535157799720764, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0176, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.10429015755653381, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0188, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.09504082053899765, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0186, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.12917572259902954, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0182, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.11084037274122238, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0163, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.12090355157852173, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0196, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.11049258708953857, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0194, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.0831870436668396, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0188, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11006250977516174, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.018, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.12107989192008972, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0178, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.10361241549253464, + "learning_rate": 2.089684155251222e-05, + "loss": 0.017, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.13428105413913727, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0198, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.13375632464885712, + "learning_rate": 2.088923894557553e-05, + "loss": 0.018, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.14561115205287933, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0181, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.09850531816482544, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0176, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.11646582931280136, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0162, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.14659325778484344, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0168, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.15971294045448303, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0198, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.11780509352684021, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0156, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11453750729560852, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0179, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.09094218909740448, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0172, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.13252301514148712, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0209, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.10068004578351974, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0159, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.10302501171827316, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0172, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.10893740504980087, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.019, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.12446500360965729, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0177, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1345142424106598, + "learning_rate": 2.083597073454921e-05, + "loss": 0.021, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.12400489300489426, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0194, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.15042100846767426, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0162, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.1559734046459198, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0189, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.10689649730920792, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0163, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.15904614329338074, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0199, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.11539188027381897, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0208, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.14550678431987762, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0191, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.10212639719247818, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0177, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.10318224877119064, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0176, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0196, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.10269734263420105, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0179, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.11070621758699417, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0196, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13299307227134705, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.02, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11430692672729492, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0188, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10718249529600143, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.0204, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.11855681240558624, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.019, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.10754020512104034, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0157, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11011523753404617, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0172, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.11931152641773224, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.0171, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.12425696104764938, + "learning_rate": 2.075972279799841e-05, + "loss": 0.0195, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.10292454063892365, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0177, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.11290059238672256, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0165, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.1453622728586197, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0207, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.12152253836393356, + "learning_rate": 2.07444521407078e-05, + "loss": 0.017, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.134286567568779, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0198, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.11180262267589569, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0196, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.12045837193727493, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.018, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.18509334325790405, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0209, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.17120368778705597, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0197, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.11996874213218689, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0204, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.12391912192106247, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0201, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.1248779147863388, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.016, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.12873846292495728, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0184, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.11072682589292526, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.0182, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.10367089509963989, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0179, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.12854290008544922, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0184, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.16348853707313538, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0169, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.1345866173505783, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0187, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.1139485239982605, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0178, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.1656051129102707, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0177, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.1133129820227623, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0154, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.0865294560790062, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0163, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.1368715763092041, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0218, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.11601361632347107, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0185, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.1401238590478897, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0177, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.10183753073215485, + "learning_rate": 2.066033925513678e-05, + "loss": 0.0192, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.10317160934209824, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0153, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.12317138910293579, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.021, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.10953490436077118, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0191, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.10530596971511841, + "learning_rate": 2.064502355956822e-05, + "loss": 0.0179, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.09054628014564514, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0189, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.12472135573625565, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0239, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.1551167368888855, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0189, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.08141466975212097, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0174, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.13025878369808197, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0185, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.13056428730487823, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0183, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.1187913790345192, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0182, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.11957527697086334, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0173, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.09339714050292969, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0182, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.1584015190601349, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.02, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.13754191994667053, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0193, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.09880727529525757, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0192, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.15097497403621674, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0172, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12028415501117706, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0182, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.10583297163248062, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0183, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.09343970566987991, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0187, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.1436399519443512, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0175, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.10790709406137466, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0184, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.14520230889320374, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0187, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.10826292634010315, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0188, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.1122521162033081, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0197, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.11445961147546768, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.015, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.1473265290260315, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0193, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.1572754830121994, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0186, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.1085842102766037, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0169, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.11982112377882004, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0163, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.12014015763998032, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.0217, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.11440391838550568, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0175, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.1276320368051529, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0203, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12434501200914383, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0186, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.11981593072414398, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0138, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.08389589190483093, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0162, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.11693917959928513, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.0199, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.1033058762550354, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0182, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.17878879606723785, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0181, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.12847506999969482, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0186, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.10514750331640244, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0166, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.1037413701415062, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0182, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.10402073711156845, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0165, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.10330315679311752, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0164, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.14082680642604828, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0205, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.1187078058719635, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0178, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.10590347647666931, + "learning_rate": 2.047995045386578e-05, + "loss": 0.0192, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.11178472638130188, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.019, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.11713241040706635, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0169, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.1520795077085495, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0173, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.11906655877828598, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0163, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.10854434967041016, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0181, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.16785599291324615, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.0212, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10305774211883545, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0179, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.09346644580364227, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0162, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11048682034015656, + "learning_rate": 2.044530211206255e-05, + "loss": 0.0158, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.14503270387649536, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0185, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.12631897628307343, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0182, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.14716260135173798, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0205, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.1364247053861618, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0187, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.09775770455598831, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0179, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.17999885976314545, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0194, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.1381615400314331, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0192, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.10749474912881851, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0189, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.15220865607261658, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.12083570659160614, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0175, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.11410468816757202, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0174, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.11285928636789322, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.0191, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.09808669239282608, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0173, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.07822877913713455, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0179, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.1032382920384407, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0177, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12453432381153107, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0183, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.1397278904914856, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0175, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.1701551228761673, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0176, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.13178907334804535, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0158, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.16403238475322723, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0182, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.13050618767738342, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.019, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09723208099603653, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.016, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.11937404423952103, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0194, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.10595990717411041, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0179, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.10598801076412201, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0172, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.12562280893325806, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0192, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.1062425896525383, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0167, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.19852802157402039, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0186, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10752324014902115, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0171, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.11313100904226303, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0191, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.09881386160850525, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0189, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.11926184594631195, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0176, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.11304130405187607, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0171, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.159687340259552, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0209, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.12324368953704834, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0184, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.10718217492103577, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0178, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12370229512453079, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0181, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.11843078583478928, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.017, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.11866020411252975, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0173, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.10999849438667297, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0217, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.10843008756637573, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.0163, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.13494807481765747, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0202, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.13634233176708221, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.0203, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.09910089522600174, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0154, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.1105995923280716, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0172, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.11397486180067062, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0171, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10141155868768692, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0187, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.09431210905313492, + "learning_rate": 2.025995044916193e-05, + "loss": 0.0163, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.09627058357000351, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0188, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.10498029738664627, + "learning_rate": 2.025220726516947e-05, + "loss": 0.0197, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.12692295014858246, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0163, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.11992908269166946, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0188, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.1448463350534439, + "learning_rate": 2.024058949838742e-05, + "loss": 0.0182, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.10764076560735703, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0191, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.1176103726029396, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0202, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.10393970459699631, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0174, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.12357908487319946, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0188, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11184806376695633, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0171, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09287009388208389, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0164, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.13145606219768524, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0175, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.108634814620018, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0192, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.1643940508365631, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.02, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.1642858386039734, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0195, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.11515326052904129, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0173, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.11438507586717606, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0187, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.1556805521249771, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0193, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.1081680878996849, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0176, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.11440131068229675, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0198, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.13953827321529388, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0163, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.12443852424621582, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0173, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.12884436547756195, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.0184, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.09170050173997879, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0158, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.08685769885778427, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0143, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.12880417704582214, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0172, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.09957600384950638, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0189, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.11359020322561264, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0191, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.10620398819446564, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0183, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.11106064170598984, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0175, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.14327868819236755, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0156, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.09743347764015198, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0165, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.11041118204593658, + "learning_rate": 2.013198534613597e-05, + "loss": 0.019, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.13757234811782837, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.017, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.10167638957500458, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0172, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.1386769413948059, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0157, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.13493874669075012, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0184, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.1238163560628891, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0178, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12875531613826752, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0176, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11199075728654861, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0179, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12402849644422531, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0197, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.12012362480163574, + "learning_rate": 2.009701166604228e-05, + "loss": 0.0203, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.1319619119167328, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0176, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.11875088512897491, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.017, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.14083963632583618, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0192, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1157609224319458, + "learning_rate": 2.008145773123833e-05, + "loss": 0.017, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1447380781173706, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0192, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.08668482303619385, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.0182, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.11417029052972794, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0162, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.12431302666664124, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0176, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.09466125816106796, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0174, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.13191643357276917, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0162, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.16682866215705872, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0167, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.09262735396623611, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0174, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10246162116527557, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0157, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.14533615112304688, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0193, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10985267907381058, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0185, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.1364467293024063, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0174, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.12792983651161194, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.019, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.13055621087551117, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0175, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.1272381991147995, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0191, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.1523309350013733, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.0192, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.12438428401947021, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0183, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.13417088985443115, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0192, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.10955093055963516, + "learning_rate": 2.000749254391884e-05, + "loss": 0.019, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.1140032708644867, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.0155, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.13144028186798096, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.0177, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.10048601031303406, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0169, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.1272922307252884, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0167, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.10301118344068527, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0166, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.10341407358646393, + "learning_rate": 1.998410649877165e-05, + "loss": 0.018, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.08479570597410202, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0193, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.10605745017528534, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0169, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.09806226938962936, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0175, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.12049946188926697, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0187, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.13633446395397186, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0167, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.12669390439987183, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.02, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.1395879089832306, + "learning_rate": 1.99568055617237e-05, + "loss": 0.02, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.1145915761590004, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.019, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12087064981460571, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0194, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.11252355575561523, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0156, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.10857690125703812, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0197, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.17054887115955353, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0172, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.12123709172010422, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0184, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.08856352418661118, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0178, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.09841718524694443, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0198, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.1236206591129303, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0171, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11047739535570145, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0198, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.11695606261491776, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0162, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.12013910710811615, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0171, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.08011548966169357, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0176, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.13469302654266357, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0177, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.12169095873832703, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.0192, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.1165102869272232, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0165, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.16436143219470978, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0164, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.11628631502389908, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0174, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.13279633224010468, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0164, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11530247330665588, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0199, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13026569783687592, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0193, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.12264525890350342, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0178, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.09320709109306335, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0157, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.11837475001811981, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.016, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.12293318659067154, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.0182, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.09163888543844223, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0197, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.1171371266245842, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0162, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.1096082329750061, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0198, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09364283084869385, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0164, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.08231621980667114, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0169, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.1273244321346283, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.0167, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.15531282126903534, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0194, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.13366153836250305, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0186, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.1072721853852272, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0173, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.11176268011331558, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0182, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.10074266046285629, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0167, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.11638782918453217, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0194, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.1109209880232811, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0197, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.1511884480714798, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0202, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.09859053790569305, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0185, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.12002220004796982, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0177, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.11930137127637863, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0159, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.10564088076353073, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0176, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.10428094118833542, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0197, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.1335987150669098, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0188, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.12865577638149261, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0179, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.14246998727321625, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0181, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.1250566691160202, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0188, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.09627831727266312, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0153, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.09008506685495377, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0185, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.12598916888237, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0169, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.11335116624832153, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0165, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.10588988661766052, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0171, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.10731689631938934, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.0174, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.11184538155794144, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0154, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.10870286822319031, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0163, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.11692065745592117, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0178, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.10036200284957886, + "learning_rate": 1.972989772246684e-05, + "loss": 0.016, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.12077154219150543, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0178, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.1272457093000412, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0188, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.08074570447206497, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0159, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.11381031572818756, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.0155, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.09836770594120026, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0161, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.09236133843660355, + "learning_rate": 1.970635454034925e-05, + "loss": 0.0183, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.1167517825961113, + "learning_rate": 1.970242942396647e-05, + "loss": 0.018, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.10953414440155029, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.015, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.12463351339101791, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0192, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.1285952925682068, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0172, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.10794857144355774, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0167, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.11815376579761505, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0176, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.10251157730817795, + "learning_rate": 1.967887123661664e-05, + "loss": 0.018, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.16941046714782715, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0182, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.14373992383480072, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.0177, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.11837214231491089, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0169, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.10599380731582642, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0167, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.11006144434213638, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0207, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.11699084937572479, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0164, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.10131341218948364, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0177, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.09975451976060867, + "learning_rate": 1.964744043284577e-05, + "loss": 0.017, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.08056963980197906, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0171, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.16518834233283997, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0186, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.10148414224386215, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0181, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.14867499470710754, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0183, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.16952896118164062, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0199, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13692231476306915, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0198, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.15133927762508392, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0184, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11434897035360336, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0189, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.0917418971657753, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0167, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12218283116817474, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0168, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.12193279713392258, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0173, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.09389539062976837, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0182, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.10455376654863358, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0169, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.11151516437530518, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0188, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.14512453973293304, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0194, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.10494859516620636, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0169, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.15839508175849915, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0177, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.12615326046943665, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0172, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.1135321632027626, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0166, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.12872013449668884, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0173, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.18429434299468994, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0165, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.13249747455120087, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0183, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.10603360831737518, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0145, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.10620120912790298, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0155, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.10986930131912231, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0163, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09298109263181686, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0179, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09945658594369888, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0178, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.08814677596092224, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0186, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.09042713046073914, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0173, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.11592321842908859, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0191, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.1026391014456749, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.018, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.09040834754705429, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0164, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.11391181498765945, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0177, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11113253980875015, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.0166, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.10178712755441666, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0167, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.13696520030498505, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0171, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.10290519148111343, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0164, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13472187519073486, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0173, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.08172356337308884, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0187, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.12297531962394714, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0166, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.13925355672836304, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0193, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.15394000709056854, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0169, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.09050899744033813, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.019, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.11854210495948792, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0181, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.10207716375589371, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0154, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.10006722062826157, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0161, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.10565274208784103, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0183, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.11656428128480911, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0166, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.12020296603441238, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0181, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.0964578241109848, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0174, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.0888240784406662, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0158, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.09271078556776047, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0187, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.10407985746860504, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0152, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1329568326473236, + "learning_rate": 1.94346993525268e-05, + "loss": 0.0166, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.13284213840961456, + "learning_rate": 1.943075029574398e-05, + "loss": 0.018, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.14158464968204498, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0194, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.11495806276798248, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0157, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.12886112928390503, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0174, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08954162895679474, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0158, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.09367296099662781, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0171, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.11188989877700806, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0162, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.10758768022060394, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0154, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.11783428490161896, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0182, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.18594273924827576, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0175, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.08010479807853699, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0167, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13949649035930634, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0175, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.10709669440984726, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0179, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.1675017923116684, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0176, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12764538824558258, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0165, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11073391884565353, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0169, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.10426445305347443, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0171, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.09000374376773834, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0161, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.10287991911172867, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0171, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.12322332710027695, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0178, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.11188322305679321, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0166, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.14056499302387238, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0196, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.08969348669052124, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0169, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.10100627690553665, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0159, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.1181580200791359, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0185, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.11281560361385345, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0171, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.11793895065784454, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0187, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.14534896612167358, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0177, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.10177207738161087, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0182, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.1335727721452713, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.017, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.14608687162399292, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0197, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.09763725847005844, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0184, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.09486788511276245, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0177, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.12235541641712189, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.018, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.13838450610637665, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0174, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.09837213158607483, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.0174, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.11713139712810516, + "learning_rate": 1.928836248385875e-05, + "loss": 0.016, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.13015268743038177, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0187, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.13498486578464508, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0181, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.12189928442239761, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0156, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.14399787783622742, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0185, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.10662715882062912, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0164, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.1145375519990921, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0185, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.11677944660186768, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0153, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.10242128372192383, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0168, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.1290024369955063, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0189, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.11820637434720993, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0167, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.12113048881292343, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0187, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.12122460454702377, + "learning_rate": 1.92408053936279e-05, + "loss": 0.0175, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.09413880109786987, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0149, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.11389796435832977, + "learning_rate": 1.923287468901755e-05, + "loss": 0.0185, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13923633098602295, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.0169, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.15947312116622925, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0188, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.14314432442188263, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0182, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.1426798403263092, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0179, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.12179753184318542, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.02, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11758018285036087, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0197, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11349253356456757, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0171, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.13724955916404724, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0181, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.1260363608598709, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0172, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.09541817009449005, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0181, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.0922083631157875, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0177, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.08051180839538574, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0167, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.11803245544433594, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0177, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10018756985664368, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.10512874275445938, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0164, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.17478078603744507, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0169, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.08864858746528625, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0165, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11313372850418091, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0172, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.09432164579629898, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0163, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.09435626119375229, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0188, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.10041314363479614, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0157, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.07982632517814636, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0176, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.09699293971061707, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0177, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.11059583723545074, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.017, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.11384644359350204, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0163, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.11473467200994492, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0165, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.11018316447734833, + "learning_rate": 1.912568530871305e-05, + "loss": 0.0179, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.10781396180391312, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0199, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.11702359467744827, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0165, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.12909658253192902, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0162, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.09839031100273132, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0152, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.1357898861169815, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0198, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.11221814155578613, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.0178, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.11431093513965607, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0176, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.11436092853546143, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0187, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.09838918596506119, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0148, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.11489785462617874, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0168, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.10876795649528503, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0169, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.1032189428806305, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0152, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.10013218969106674, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.11554815620183945, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0182, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.2529490292072296, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0198, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.09215852618217468, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0168, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.12875677645206451, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0185, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.1232401430606842, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0179, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.10334797203540802, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0171, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10524798929691315, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0162, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.11014945805072784, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0176, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.1074177473783493, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0162, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.14043818414211273, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0191, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.10011711716651917, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0178, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0172, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.12526027858257294, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0137, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.15774008631706238, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0153, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.1049765795469284, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0155, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09307108074426651, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0186, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1369069367647171, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0158, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.12320058792829514, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0181, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.10885081440210342, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0168, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.08989177644252777, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0153, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.09219451993703842, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0172, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.11661440879106522, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0192, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.0971342995762825, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0178, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.10453730076551437, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0165, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.17652514576911926, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0196, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10733854025602341, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0154, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.0989762544631958, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0181, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.12107723951339722, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.1385297179222107, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0191, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.12209209054708481, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0217, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.09332361072301865, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0158, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.12448447197675705, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0164, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.12061425298452377, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0169, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.12132656574249268, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0183, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.14007116854190826, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0167, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.11448334902524948, + "learning_rate": 1.893057716200171e-05, + "loss": 0.015, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.12702956795692444, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0152, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.09236849099397659, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0177, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.09818986803293228, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.0164, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12131274491548538, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0171, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10609138011932373, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.0147, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.1431477814912796, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0173, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.0844314768910408, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0162, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.09800038486719131, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0157, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.11529140919446945, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0172, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.086520716547966, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0154, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.12068255245685577, + "learning_rate": 1.888667760339568e-05, + "loss": 0.015, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.12278824299573898, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0174, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.10242115706205368, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0173, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.11696609109640121, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0187, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.11133866757154465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0151, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.1319858729839325, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0165, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.08355733752250671, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0166, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.10663259029388428, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0177, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.09787473827600479, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0154, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.12111647427082062, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0171, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.10036861151456833, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.15351398289203644, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.019, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.11561891436576843, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0174, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.09197179228067398, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0174, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.0961032509803772, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.019, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.09264974296092987, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0158, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.11056753247976303, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0159, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.10464371740818024, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0165, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.13888190686702728, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0185, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.09801805019378662, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0189, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.15648017823696136, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0167, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1129915863275528, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0166, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.10456246137619019, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0171, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.10873697698116302, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.0168, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.17470459640026093, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0169, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.15504027903079987, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.10821744054555893, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0152, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.14031334221363068, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0181, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.128667414188385, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0183, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.09095419198274612, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0151, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.09303424507379532, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0167, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.10706666111946106, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0174, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.13011431694030762, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0177, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.13821059465408325, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0154, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.10680826753377914, + "learning_rate": 1.875076403503638e-05, + "loss": 0.017, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.12847846746444702, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0146, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.1472046673297882, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0184, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10622651875019073, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0153, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.129075825214386, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0189, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.12048091739416122, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.017, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.10954522341489792, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0175, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.10811172425746918, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0173, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.13378627598285675, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0193, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.13185591995716095, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0195, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.1013568863272667, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0182, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.07558644562959671, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0173, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.10668148100376129, + "learning_rate": 1.870271530813327e-05, + "loss": 0.018, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.134077250957489, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0157, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.14325739443302155, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0221, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.1389995515346527, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0178, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.13043907284736633, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0169, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.12082460522651672, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0174, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11521385610103607, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0165, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.11550573259592056, + "learning_rate": 1.867466815081593e-05, + "loss": 0.017, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.11315014213323593, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0173, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.1067388728260994, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0161, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.0941673144698143, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0158, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.09685055166482925, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0161, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.11744765937328339, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0164, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.10552993416786194, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0154, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.12654545903205872, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0184, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.11333387345075607, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0186, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.09856864809989929, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.0187, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.0933985561132431, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0171, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.096091628074646, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0194, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.08463368564844131, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0165, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.10648614913225174, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.017, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.10691184550523758, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0165, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1005169153213501, + "learning_rate": 1.861452118736741e-05, + "loss": 0.0174, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.1037023589015007, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0172, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.1299487054347992, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.015, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14639562368392944, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0167, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.12270855903625488, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0173, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.09706515073776245, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0169, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.11888926476240158, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.0152, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.11289773136377335, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0175, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13422279059886932, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0168, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.10992032289505005, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0176, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.10719195753335953, + "learning_rate": 1.85743888359202e-05, + "loss": 0.017, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.10348568856716156, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0166, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.10436776280403137, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0172, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.09779159724712372, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.018, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.10799746960401535, + "learning_rate": 1.855832827763672e-05, + "loss": 0.016, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.11972185224294662, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0195, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.10282087326049805, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0169, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.11840159446001053, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0183, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.13861460983753204, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0205, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10774466395378113, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0163, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.14446309208869934, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0168, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.1313662976026535, + "learning_rate": 1.853021190846946e-05, + "loss": 0.0161, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.0857798308134079, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0159, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.11933837085962296, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0178, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11762946099042892, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0154, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.11128528416156769, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0198, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.07360108196735382, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0148, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.10680225491523743, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0174, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.11387799680233002, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0188, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.2774786651134491, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0165, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.12314122170209885, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.017, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.14403507113456726, + "learning_rate": 1.849002290456207e-05, + "loss": 0.0175, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.10811382532119751, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0173, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.1399172991514206, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0154, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.10577643662691116, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0177, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.10794928669929504, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0155, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.0844842791557312, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0149, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.09744124114513397, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0166, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.14749610424041748, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0179, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.09801949560642242, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0177, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.07795082777738571, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0174, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.11197780072689056, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.0172, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.11819472163915634, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.08461402356624603, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0171, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.11102861911058426, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0156, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.11057187616825104, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0189, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.11031147837638855, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0155, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.08435497432947159, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0187, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.12450679391622543, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.017, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.09348208457231522, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0171, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.10584745556116104, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0161, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.14194349944591522, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.0177, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.1025288999080658, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0153, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.07800278812646866, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0157, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.1253945231437683, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0176, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.11724173277616501, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.016, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.13043718039989471, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0179, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.08728159964084625, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0177, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.1419975906610489, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0185, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.12774096429347992, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0187, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.10763564705848694, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.0163, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.12671130895614624, + "learning_rate": 1.836929808825424e-05, + "loss": 0.0177, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.10261370241641998, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0145, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12106029689311981, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.0157, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.12428469955921173, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0194, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.08797722309827805, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0175, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.14663904905319214, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.018, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.12288253754377365, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0176, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.11406450718641281, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0164, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10584931075572968, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.017, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1212797686457634, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0176, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.10806597024202347, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0159, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.10098059475421906, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0186, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.14030125737190247, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0204, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.09907040745019913, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0158, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1299905776977539, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0198, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.134843647480011, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0157, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.1048227846622467, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0154, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.12262892723083496, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0193, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.12806116044521332, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0167, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.12875764071941376, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.0149, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.09743807464838028, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0149, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10791302472352982, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0155, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10845673829317093, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0187, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.1366390436887741, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0161, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.101824089884758, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0174, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.12938271462917328, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0159, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.13503864407539368, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0196, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10241002589464188, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0153, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.10051950812339783, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.0159, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11316224932670593, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0166, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.10802372545003891, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0167, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.1365254521369934, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0176, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.11756277829408646, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.016, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13363559544086456, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0167, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.13138234615325928, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0194, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.1519714891910553, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0147, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.12152288109064102, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.017, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.0976492166519165, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0174, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.09356466680765152, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0158, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.09382101893424988, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0175, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.12188217788934708, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0163, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.09245157241821289, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0183, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.1661074161529541, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0202, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.154639333486557, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0143, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.117487333714962, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.017, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.09468945115804672, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0171, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.09632660448551178, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0147, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.12390028685331345, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0177, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.11924377828836441, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0154, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.13191300630569458, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0154, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.07403766363859177, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0171, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.10911701619625092, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0159, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.11888141930103302, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0189, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.09766717255115509, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0182, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.08285287767648697, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0145, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.13635197281837463, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0166, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.11627981066703796, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.10899969190359116, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0174, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.10002870857715607, + "learning_rate": 1.813525107780979e-05, + "loss": 0.016, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09963744133710861, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0163, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.2032729536294937, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0188, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.10394410789012909, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0178, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.1004093736410141, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0181, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.08776404708623886, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.0188, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.10926580429077148, + "learning_rate": 1.811099258182467e-05, + "loss": 0.018, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.09496597945690155, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0153, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.12430521845817566, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0152, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.11681065708398819, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0156, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.10450825840234756, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0162, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10932883620262146, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0163, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.1358717978000641, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0152, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.11744217574596405, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0188, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.14966122806072235, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0179, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.113619364798069, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.08801224827766418, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0168, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.09556800872087479, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0171, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.11116721481084824, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0171, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.10650353878736496, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0158, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.09222083538770676, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0154, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.10958139598369598, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.0189, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11020398885011673, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0166, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.08026755601167679, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0159, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.09629099816083908, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0162, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11310935020446777, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0192, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.1281706690788269, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0191, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12402708828449249, + "learning_rate": 1.802602130181036e-05, + "loss": 0.0159, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.09760920703411102, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0151, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.12342818826436996, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.018, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09423290938138962, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0163, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.10432913154363632, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0172, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.09628765285015106, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0169, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.11590366810560226, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0167, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10148236900568008, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0155, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.10599179565906525, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0155, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.11854606866836548, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.10626410692930222, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0185, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.12172172963619232, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0179, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.10842256247997284, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0173, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.11806859076023102, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0163, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11489225924015045, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0178, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.11859364807605743, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0164, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.18860413134098053, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.0168, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.13375963270664215, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0168, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.13163933157920837, + "learning_rate": 1.795310800633073e-05, + "loss": 0.0151, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.11135797202587128, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0175, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.12392686307430267, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0161, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.0995136946439743, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0165, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.10128366202116013, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0161, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.13080070912837982, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0169, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.10262267291545868, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0176, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.10983351618051529, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0163, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.09834928810596466, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0158, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.11170317232608795, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0166, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.14308896660804749, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.09110505878925323, + "learning_rate": 1.790851399590954e-05, + "loss": 0.017, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.12270113825798035, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0165, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.1422351896762848, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0177, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1294422596693039, + "learning_rate": 1.789634733729274e-05, + "loss": 0.017, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.1157895103096962, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0192, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.12473416328430176, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0156, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.1765025109052658, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0161, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11862069368362427, + "learning_rate": 1.78801220478228e-05, + "loss": 0.017, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.10549142211675644, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0165, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.09818614274263382, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0171, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.13607585430145264, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0168, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.09340739995241165, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0158, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.0805685818195343, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.0207, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.11172603815793991, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0164, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.08757263422012329, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0179, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.09590335935354233, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0199, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.1186867505311966, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.0146, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.09467568248510361, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0175, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11578786373138428, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0179, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09604159742593765, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0173, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.11337848007678986, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.0156, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.11909373849630356, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0188, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.08804072439670563, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0175, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.13454975187778473, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.0184, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.12201394885778427, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0148, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.11946159601211548, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0153, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12322071939706802, + "learning_rate": 1.78030044678245e-05, + "loss": 0.0181, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.10444212704896927, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0159, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.10723353922367096, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0158, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.10750993341207504, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0167, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.1207774356007576, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0207, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.12730205059051514, + "learning_rate": 1.778269750931628e-05, + "loss": 0.017, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.08926276862621307, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0151, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.10322589427232742, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0174, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10034044831991196, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.0184, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.12131599336862564, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0183, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.10715801268815994, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0156, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.14093570411205292, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0169, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.124595507979393, + "learning_rate": 1.775425890028875e-05, + "loss": 0.017, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.13153409957885742, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0189, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.1348494291305542, + "learning_rate": 1.77461316962853e-05, + "loss": 0.019, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.10926292091608047, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0161, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.11094044893980026, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0155, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.1341349333524704, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0184, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.15732111036777496, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0203, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.07597270607948303, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.0184, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.11097310483455658, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.0176, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.12087065726518631, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0164, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12998445332050323, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0191, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.13928557932376862, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0157, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.10487896203994751, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0172, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.13437892496585846, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0182, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.20842674374580383, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0189, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.09813268482685089, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12833304703235626, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0168, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.1019122451543808, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0159, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.07276969403028488, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.016, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09712188690900803, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0169, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.10403666645288467, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0156, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.11877460777759552, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0152, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.10731421411037445, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0188, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.12387207895517349, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0173, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.10561594367027283, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0179, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.2772648334503174, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0168, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.097810760140419, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0142, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.1168239563703537, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0171, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.14195922017097473, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0162, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11385945230722427, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0187, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.1124262586236, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.13074515759944916, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0151, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.12591321766376495, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0177, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.09103676676750183, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0169, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.12457307428121567, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0162, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.11226223409175873, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0156, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.14778819680213928, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0187, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.10433483868837357, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0187, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.08798553794622421, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0163, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.09729046374559402, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0149, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10188018530607224, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0154, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.12559989094734192, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.018, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12567821145057678, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0168, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.12211210280656815, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0166, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.11825426667928696, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0168, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.11676980555057526, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0175, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.11447525024414062, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0162, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.08958559483289719, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0161, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.11528483778238297, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0162, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.08296159654855728, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.014, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.12306056916713715, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0158, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.13315193355083466, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.0164, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.09533339738845825, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0166, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.20049314200878143, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.017, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.14355438947677612, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0187, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.11443490535020828, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0146, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.10336155444383621, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0164, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11926902830600739, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0173, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.09699229151010513, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0176, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.11406200379133224, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0164, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.10817506164312363, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0165, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.11398246139287949, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.0163, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.09427376091480255, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0166, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.10372194647789001, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0148, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.09003781527280807, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0172, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.09807098656892776, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.11895686388015747, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0163, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10839524865150452, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0173, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.20792271196842194, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0183, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.07746267318725586, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0147, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.11470094323158264, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0177, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.1377800703048706, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.1247936338186264, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.016, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.1405080407857895, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0164, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.1318410336971283, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0175, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.14254692196846008, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0159, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.12803269922733307, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0146, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11180103570222855, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0169, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.16513720154762268, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0162, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.13401633501052856, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.0182, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.13510771095752716, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0175, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.12391090393066406, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0174, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1019999086856842, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0157, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.11906205862760544, + "learning_rate": 1.7416306237948e-05, + "loss": 0.0179, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11846558749675751, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.0154, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.12292807549238205, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0163, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10195571929216385, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0186, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.1032365933060646, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0163, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1241045668721199, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0173, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.10956405848264694, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.0163, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.12094716727733612, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.018, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.1353612244129181, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0161, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.0935104638338089, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0164, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.10712502896785736, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.0177, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.12715180218219757, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0148, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.09601753205060959, + "learning_rate": 1.736733684019633e-05, + "loss": 0.0167, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.11189524084329605, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0174, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.11025495082139969, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.016, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.1203494742512703, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0192, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.12150877714157104, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0143, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.11957377195358276, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0146, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.16482087969779968, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0181, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.10647732019424438, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0178, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.13345423340797424, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0155, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.13268540799617767, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0178, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.12209953367710114, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0161, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.11951886862516403, + "learning_rate": 1.73224254699117e-05, + "loss": 0.0213, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.12044327706098557, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0195, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.10611505806446075, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0169, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.12347753345966339, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0145, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13930706679821014, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0185, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.12702012062072754, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0167, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.1193593367934227, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0169, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.16778719425201416, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0179, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11699119210243225, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0159, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11571445316076279, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.08481728285551071, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0159, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.10007977485656738, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0155, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.12851299345493317, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0167, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.13929468393325806, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0154, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.11237664520740509, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0143, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.11654951423406601, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0179, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.11065302044153214, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0168, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.10024461150169373, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0173, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.1334248036146164, + "learning_rate": 1.724888829234145e-05, + "loss": 0.016, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.1115458682179451, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0159, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.15415695309638977, + "learning_rate": 1.72407140459686e-05, + "loss": 0.0166, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10965333878993988, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0159, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.10152038931846619, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0152, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.10030021518468857, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0161, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.1526176780462265, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0134, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13154515624046326, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.017, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.12260743975639343, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0152, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.07395336031913757, + "learning_rate": 1.721209883968656e-05, + "loss": 0.0155, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.1107548177242279, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0178, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.08634423464536667, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0148, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.10544566810131073, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0151, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.13896141946315765, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0171, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10351476818323135, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0164, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.11912661790847778, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0163, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.12101101130247116, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0185, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.09519081562757492, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0137, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.13252407312393188, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0153, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.08964477479457855, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0148, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.11772924661636353, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0153, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1261531263589859, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.0155, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.12295454740524292, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0162, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.14834828674793243, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0162, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.14458344876766205, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0146, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.10006255656480789, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0183, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11527495086193085, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0174, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.11176037788391113, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0146, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.11118462681770325, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0139, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.1216764822602272, + "learning_rate": 1.713029611235338e-05, + "loss": 0.018, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10572952032089233, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0155, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.09299932420253754, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0193, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.11654805392026901, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0146, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.125601127743721, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0148, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.12143341451883316, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0144, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.10944563895463943, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0188, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.10531438142061234, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.017, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.08527344465255737, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0134, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.12489438056945801, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0184, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.12359192967414856, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0162, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.09504237025976181, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0147, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.1144440770149231, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0155, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.11942129582166672, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.0161, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.15496809780597687, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0167, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.11641886830329895, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0155, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.13092041015625, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0164, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.1662251204252243, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0169, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.1151951253414154, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0146, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.13008327782154083, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.017, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.08816810697317123, + "learning_rate": 1.70484286724112e-05, + "loss": 0.0147, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.12602686882019043, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0167, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10360333323478699, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0161, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.10288142412900925, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0174, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.1306314915418625, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0178, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.11859587579965591, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0152, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.12666994333267212, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.0155, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.10859639197587967, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0174, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.1423543244600296, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0145, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.12444380670785904, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0165, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.07491649687290192, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0174, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.12173314392566681, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0156, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.11790800094604492, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.1375807374715805, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0182, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.11894135922193527, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0149, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.09377862513065338, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0157, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1448003053665161, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0189, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.14296171069145203, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.019, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.10862060636281967, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0188, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.1081303209066391, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.0166, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08262617141008377, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0163, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12013579159975052, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0161, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.06014449894428253, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0166, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09688957780599594, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0164, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.11730696260929108, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.0142, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.11243025213479996, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0134, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11974845081567764, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0156, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.1408282220363617, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0193, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.1495579332113266, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.017, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.1408172994852066, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.19624975323677063, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0165, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09940195083618164, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0168, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.11277889460325241, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0171, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.13027498126029968, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0165, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.11570801585912704, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.0159, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.12621626257896423, + "learning_rate": 1.690501241902545e-05, + "loss": 0.016, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.13131381571292877, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0146, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.11577112972736359, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0168, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.1240323930978775, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0162, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.17400938272476196, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.0155, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.09539356827735901, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0165, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10108646005392075, + "learning_rate": 1.688040861067922e-05, + "loss": 0.016, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.1295158863067627, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0151, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.13638637959957123, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0188, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.12255847454071045, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0148, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.13242852687835693, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0166, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11082842200994492, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.015, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.14538930356502533, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0172, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1609591394662857, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0166, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.10611626505851746, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.0142, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.13291358947753906, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0179, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.1112443208694458, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0179, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.10251478105783463, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0174, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.10443990677595139, + "learning_rate": 1.683118563837068e-05, + "loss": 0.016, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.08651736378669739, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0176, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.09996092319488525, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.017, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.13269121944904327, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.0154, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.121453657746315, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0167, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.10587169975042343, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0161, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.12990498542785645, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0158, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.12159401923418045, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0158, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.12046501785516739, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0165, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.10626068711280823, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0144, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.14602169394493103, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0161, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09537681937217712, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0188, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.1014745756983757, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0171, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.11358807235956192, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0165, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.12290225178003311, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0153, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.11178604513406754, + "learning_rate": 1.676962882515376e-05, + "loss": 0.0175, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.1464385688304901, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0155, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.12778854370117188, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0144, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.09556689858436584, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0171, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.10866247862577438, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.0151, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.14120467007160187, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0163, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.12688933312892914, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0174, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.10251957923173904, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0171, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.11465741693973541, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.016, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.08181405812501907, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0201, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.09379345923662186, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0172, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.12122894823551178, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0183, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.0956249088048935, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0142, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.11953603476285934, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0169, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.12494538724422455, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0167, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.20808497071266174, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0138, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.10776039958000183, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0152, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.09606330841779709, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0154, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.10988017171621323, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0182, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.10328707098960876, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0146, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.166867196559906, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0179, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.11701484769582748, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0159, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.10510280728340149, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0179, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.12857040762901306, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0172, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1627577394247055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0176, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.10962127149105072, + "learning_rate": 1.666696746039845e-05, + "loss": 0.0147, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.12961529195308685, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0164, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.08055070787668228, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0148, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.10041354596614838, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0161, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.12196766585111618, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0138, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.0879533439874649, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0156, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.1102190688252449, + "learning_rate": 1.664231677578745e-05, + "loss": 0.016, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.12463485449552536, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.016, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.10829959809780121, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0181, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.08602527529001236, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0176, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.10241527855396271, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.0172, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11007986962795258, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0156, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13110534846782684, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.018, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1417531818151474, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0161, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.13233700394630432, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0147, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.09148544073104858, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0145, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.08781567215919495, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.0149, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.10577751696109772, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0161, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.11689155548810959, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0171, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.0777219608426094, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0152, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.07983718812465668, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0158, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1595761626958847, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0168, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.1037725880742073, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0161, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.10239970684051514, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0149, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.15451422333717346, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0199, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.11882807314395905, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0165, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.09850180149078369, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0147, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.10899021476507187, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.09254994243383408, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0148, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.13559140264987946, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0162, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.11541012674570084, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0154, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.07937203347682953, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0153, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.10062690824270248, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0165, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.13341091573238373, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0178, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.0857294499874115, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0132, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.10551585257053375, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0142, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.12363193184137344, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0165, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.11550596356391907, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0165, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.09379533678293228, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0154, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.10079154372215271, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0163, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.13103018701076508, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0178, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.1172979474067688, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0152, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.11038312315940857, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0152, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.0925338938832283, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0148, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.2519476115703583, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.017, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.09703299403190613, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0193, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.10749004036188126, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0175, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.08736859261989594, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0147, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.13125599920749664, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0161, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10677823424339294, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0156, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.09135954827070236, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0151, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.10548236966133118, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0179, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.1150064468383789, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0188, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.127374529838562, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.016, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.09152884781360626, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0177, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.1021062359213829, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0145, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.1054648831486702, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0172, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.12128820270299911, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0165, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.11375363171100616, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0151, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.11004973948001862, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0182, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.11101444810628891, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0164, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.10393026471138, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0172, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.1139075756072998, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0159, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1569976508617401, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0178, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09649650752544403, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0165, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.16198694705963135, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0157, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.10348767787218094, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0173, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.12259576469659805, + "learning_rate": 1.639145830609088e-05, + "loss": 0.0171, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.10480271279811859, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0159, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.11254257708787918, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0163, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.11534930765628815, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0149, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.10350872576236725, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0169, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.10812029242515564, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.015, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.11796360462903976, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.0144, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.1276526153087616, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0157, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.11665219813585281, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0151, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.10940699279308319, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0165, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.09596745669841766, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0149, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15527229011058807, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0164, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11651212722063065, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0145, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.12004360556602478, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0153, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.13298730552196503, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0176, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.1496417224407196, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0153, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.11104541271924973, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0158, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.12215209007263184, + "learning_rate": 1.632147383062092e-05, + "loss": 0.015, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.09335075318813324, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0148, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.11481857299804688, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0153, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14369069039821625, + "learning_rate": 1.630912057457699e-05, + "loss": 0.019, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.11009083688259125, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0162, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.11112431436777115, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.0137, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.10135089606046677, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0167, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09966760873794556, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0168, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12062226980924606, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0145, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.1109282448887825, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0162, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.10862662643194199, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0152, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.09910363703966141, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0156, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.1374131590127945, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0167, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.12114953994750977, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.0156, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.14391179382801056, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0156, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.12364984303712845, + "learning_rate": 1.625969868712804e-05, + "loss": 0.0169, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.08771048486232758, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0142, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.10959397256374359, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0176, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.138074591755867, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0149, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.1116161122918129, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0154, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.1398090422153473, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0164, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.12145344167947769, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0173, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.11162104457616806, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0153, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.1211397722363472, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.016, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09488996863365173, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0146, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.11799386143684387, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0137, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.1406150758266449, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.0149, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.1307811290025711, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0192, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.13287024199962616, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.0153, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10396360605955124, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0133, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.11891327053308487, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0167, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.12922705709934235, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0178, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10563277453184128, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0132, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.17530488967895508, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0159, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.10669656842947006, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0166, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.09745346754789352, + "learning_rate": 1.617729852865125e-05, + "loss": 0.0152, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.12519428133964539, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0136, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.10747594386339188, + "learning_rate": 1.616905650452066e-05, + "loss": 0.0147, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.10352121293544769, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0161, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.1046888679265976, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0131, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.2108439952135086, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0148, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.1401670128107071, + "learning_rate": 1.61525713933778e-05, + "loss": 0.0171, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.10765372961759567, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0169, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.14557960629463196, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0151, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.11173100024461746, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0181, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.08689062297344208, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0156, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.10052058845758438, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0171, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.11383800208568573, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0167, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10390642285346985, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0157, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.10242070257663727, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0147, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.10699810087680817, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.0162, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.1349048614501953, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0167, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.10787097364664078, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.0144, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.06869964301586151, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0168, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.08449438959360123, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.0156, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10576694458723068, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0157, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.10042312741279602, + "learning_rate": 1.609073991709496e-05, + "loss": 0.018, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.10223086178302765, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0179, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.11301153153181076, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0151, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.11257045716047287, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0159, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.09129524976015091, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0154, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.10601308941841125, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0157, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.11378258466720581, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0151, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.1138806939125061, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0188, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.10367786139249802, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0157, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.13393838703632355, + "learning_rate": 1.605363201758077e-05, + "loss": 0.016, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.09532077610492706, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0145, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.13368268311023712, + "learning_rate": 1.60453849294712e-05, + "loss": 0.0178, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10675307363271713, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0148, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.10722523182630539, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0167, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.09825004637241364, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0148, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.14284753799438477, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.0194, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0167, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.13374702632427216, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0153, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.08964364975690842, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.0164, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.07354595512151718, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0138, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.08927512913942337, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0167, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.10888224840164185, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0193, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.12273052334785461, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0157, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.10807470232248306, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0152, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.1124584972858429, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.0148, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.10052672028541565, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0145, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.08666819334030151, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0172, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.11364969611167908, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0155, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.11263672262430191, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0164, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.1379120796918869, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0187, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.10857756435871124, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0133, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.13408084213733673, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0156, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.10847405344247818, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0141, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.10151072591543198, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0138, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11481117457151413, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0138, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.10202853381633759, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0153, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.09370319545269012, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.0159, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.09684473276138306, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0162, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.11064363270998001, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0162, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.15720140933990479, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0167, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.12132807075977325, + "learning_rate": 1.592576773310894e-05, + "loss": 0.0159, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.10888595134019852, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0133, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.0892426148056984, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0163, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12278196215629578, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0186, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10831111669540405, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0155, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.08722463995218277, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0156, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.11901626735925674, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0155, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.08866825699806213, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.0155, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.09303046017885208, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0152, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.12453419715166092, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0166, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.11046329140663147, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0155, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.10482461750507355, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0154, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11235640943050385, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0165, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.11749114841222763, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0159, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.1201302632689476, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0174, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.13044941425323486, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0143, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.10600301623344421, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0155, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.12556208670139313, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0159, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.12454571574926376, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0153, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.13055634498596191, + "learning_rate": 1.584736531020125e-05, + "loss": 0.0188, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.1379219889640808, + "learning_rate": 1.584323820456826e-05, + "loss": 0.016, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.13937926292419434, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0181, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.21694405376911163, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.013, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.11387006938457489, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0144, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.10102641582489014, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0154, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.10343687981367111, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0161, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.14018288254737854, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0175, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10792441666126251, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0168, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.10692388564348221, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0175, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.11248136311769485, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0176, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.13729289174079895, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0154, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.1389625370502472, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0152, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11402085423469543, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0154, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.14007720351219177, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0151, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.10529543459415436, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0151, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.09746415913105011, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0136, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1150866374373436, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0158, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.10207786411046982, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0139, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.13951019942760468, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0157, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11709082126617432, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.0158, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.12401671707630157, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0145, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.103446826338768, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0149, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.16494140028953552, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0192, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.10893384367227554, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0141, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.09813039004802704, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.0142, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.14741374552249908, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0184, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.12035968154668808, + "learning_rate": 1.573591189751e-05, + "loss": 0.0157, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.0970459058880806, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0176, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.10946980863809586, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0144, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.15628819167613983, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0157, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.12485401332378387, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0184, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.11361467838287354, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0163, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.19872187077999115, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0152, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.0912356749176979, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0167, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.09560416638851166, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0178, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.10187338292598724, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0127, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.11934805661439896, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0145, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.11719757318496704, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0148, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09480439126491547, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0167, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.09031171351671219, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0163, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.12236128002405167, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.014, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.09773887693881989, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.015, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.11999719589948654, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.0145, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.11341046541929245, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0142, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.10415472090244293, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0183, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.10177286714315414, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0152, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.10778342932462692, + "learning_rate": 1.565332720308093e-05, + "loss": 0.016, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.08542678505182266, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0157, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.10727313905954361, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0164, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.1413055658340454, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0163, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.10777363926172256, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0175, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.14286576211452484, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0155, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.10413800925016403, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0164, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.10655509680509567, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0163, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.10989648103713989, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0153, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.09240605682134628, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0144, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.0842762365937233, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0143, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.13544097542762756, + "learning_rate": 1.560789700123175e-05, + "loss": 0.017, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11756332218647003, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0159, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.11726696044206619, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0165, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.11693235486745834, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0146, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.1371903419494629, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0145, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.0971926674246788, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.017, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.08220933377742767, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0156, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.10194069892168045, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0134, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.08653450012207031, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.0131, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.08990923315286636, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0144, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.10370788723230362, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0133, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.06916087865829468, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.0139, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.1136801689863205, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0146, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12992094457149506, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.0144, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.09605134278535843, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0165, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.10273651778697968, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0165, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.11571857333183289, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0151, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.10102164000272751, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0146, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.11045269668102264, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.0146, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.12626338005065918, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0166, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.1357768476009369, + "learning_rate": 1.552528261458322e-05, + "loss": 0.018, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.10587792843580246, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0143, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.1509760171175003, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0187, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11034028977155685, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0174, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.07888679206371307, + "learning_rate": 1.550875774213821e-05, + "loss": 0.0159, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.11219298094511032, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.13006654381752014, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.0161, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12485881894826889, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0148, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.10320480912923813, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0171, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.12367358803749084, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.019, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1412898451089859, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0196, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.10843265801668167, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.0146, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09489759802818298, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0147, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.10236489772796631, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0158, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.12652312219142914, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0173, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10095653682947159, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0167, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.0802559182047844, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0145, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.1239897757768631, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0155, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.09682597219944, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.0154, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.10482888668775558, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.013, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.1272023320198059, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0164, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.09988776594400406, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0166, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.10352584719657898, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0142, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.36612969636917114, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0169, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.12097200751304626, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09437882900238037, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.0147, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14130210876464844, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0147, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.1284940391778946, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0148, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12240798771381378, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0135, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.08258320391178131, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0154, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.10075270384550095, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0162, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.10424498468637466, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0138, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.10447008907794952, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0156, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.10173123329877853, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0127, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.12066177278757095, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0162, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.09489274024963379, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0159, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.09174267202615738, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0155, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.13935479521751404, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0158, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.11126208305358887, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0142, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.13917426764965057, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.014, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.10231155157089233, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0141, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.11477960646152496, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0145, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.10886958241462708, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0155, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.09005019068717957, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0147, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.09915502369403839, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0138, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.1025547906756401, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0158, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.1030225157737732, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0158, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.0842776820063591, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0163, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.0634579285979271, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0151, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.09493366628885269, + "learning_rate": 1.532281505788039e-05, + "loss": 0.014, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12343619018793106, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0167, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.10681836307048798, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0149, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.1302853226661682, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0171, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.10506479442119598, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0136, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.12051218003034592, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0168, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11805758625268936, + "learning_rate": 1.52980183248759e-05, + "loss": 0.0158, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.10309842228889465, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.015, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.108099564909935, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0159, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11023534834384918, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0136, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.11571311205625534, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0163, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.1068359762430191, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.017, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.15622149407863617, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0156, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.11074791103601456, + "learning_rate": 1.526908777801797e-05, + "loss": 0.0163, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.11449336260557175, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0156, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.1219671219587326, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.0167, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.08449103683233261, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0162, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.13202930986881256, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0167, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.11665860563516617, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0173, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.13584332168102264, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0148, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.09913904964923859, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0155, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09704537689685822, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0162, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.13063789904117584, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0153, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.1202530562877655, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0182, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.08398830890655518, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.015, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.09059721976518631, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0129, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.10968997329473495, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0141, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.09898033738136292, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.1568402498960495, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0161, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.09885324537754059, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0168, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09452448785305023, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0142, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.16158591210842133, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0149, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.13292880356311798, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.016, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.17052333056926727, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0165, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.11351195722818375, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0163, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.09296253323554993, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0128, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.09396881610155106, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0157, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.13142332434654236, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0161, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.1246478259563446, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0135, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.13347554206848145, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0173, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.10838479548692703, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.11035843938589096, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0158, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.10406306385993958, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0149, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.16070416569709778, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0149, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09154598414897919, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.014, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.10438812524080276, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0164, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.09018818289041519, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0149, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1320919543504715, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0151, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.0870271846652031, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0155, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.12103381007909775, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0191, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.09007984399795532, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.0141, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.10315215587615967, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.19641031324863434, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0169, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14303788542747498, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0207, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.10933230072259903, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0136, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.11214675009250641, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0175, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.09008116275072098, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0145, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.09065063297748566, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.016, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.1116153672337532, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.016, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.09231071174144745, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0148, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12521710991859436, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0159, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.12240038067102432, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0143, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.09896460920572281, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0154, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.10471562296152115, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0152, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.11139113456010818, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0132, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.09953293949365616, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.015, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.12128512561321259, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0155, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.14064094424247742, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0132, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.09164810925722122, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0144, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.11127574741840363, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0145, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.11218927800655365, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0168, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.10342293232679367, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0137, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.10499373823404312, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0151, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.08029376715421677, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0157, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09445271641016006, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0161, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.09650483727455139, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0155, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.08300688117742538, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0176, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.11890654265880585, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0145, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10451909154653549, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0195, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.08795508742332458, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0141, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.1659373790025711, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0197, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.1001206561923027, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0161, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10190606862306595, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0161, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.11775147914886475, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0139, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.11382468789815903, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0149, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10219351947307587, + "learning_rate": 1.497147766283292e-05, + "loss": 0.0143, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.0930316224694252, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0148, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.10227128118276596, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0153, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.10634683817625046, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0163, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.1443759799003601, + "learning_rate": 1.495494301522528e-05, + "loss": 0.016, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.10867374390363693, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0147, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.11999388039112091, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0157, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10372292995452881, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0161, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.14905284345149994, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.0172, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10412498563528061, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0145, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.09797366708517075, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0133, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.1329793930053711, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0179, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.09716201573610306, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0146, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.09868425875902176, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.0153, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.0929884985089302, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0159, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.1417207568883896, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0151, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.13194335997104645, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0177, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.11850415170192719, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0173, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08122166991233826, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0149, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.11195199191570282, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.12405557930469513, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0135, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.12344473600387573, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0153, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.11776174604892731, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.0154, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.11634577810764313, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0158, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.11508087068796158, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0165, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.12009026855230331, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0141, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.15844415128231049, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.014, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.11803163588047028, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.014, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.11727108061313629, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.0157, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.09923852235078812, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0153, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.14936032891273499, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0174, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.10241764038801193, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0164, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.11330186575651169, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.016, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.14860020577907562, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0152, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.1636815220117569, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.015, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.11093862354755402, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0142, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.11333172768354416, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0141, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.10221018642187119, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0155, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.11085604131221771, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0158, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.17324623465538025, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0153, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.10631918907165527, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.0136, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.09826772660017014, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0148, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.1288551241159439, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11162615567445755, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0148, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.1307016909122467, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.017, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11215298622846603, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0154, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.09471828490495682, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0145, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10456835478544235, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.0147, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.08179501444101334, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0129, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.11810965836048126, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0151, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.0977158397436142, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0139, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.09650871902704239, + "learning_rate": 1.476067039948277e-05, + "loss": 0.0138, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.10262744128704071, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0145, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11392423510551453, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0147, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11900563538074493, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0163, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.10666237771511078, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0162, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.10767115652561188, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0152, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.1232513040304184, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0144, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.12271980196237564, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0167, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.12620051205158234, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0176, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11799243837594986, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0153, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.11633779853582382, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0154, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.10304611176252365, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0173, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.11186012625694275, + "learning_rate": 1.471107401829042e-05, + "loss": 0.0143, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.12054259330034256, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0174, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.1436678171157837, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0146, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.12709002196788788, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0171, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.11617328971624374, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0146, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.1117766946554184, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0178, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.12944459915161133, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0155, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11261707544326782, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0144, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.11154191195964813, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.013, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.11878378689289093, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0142, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.12532871961593628, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0176, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.10492552816867828, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0165, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.10776105523109436, + "learning_rate": 1.466148079674631e-05, + "loss": 0.016, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.10712877660989761, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0165, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.091713547706604, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0155, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.10679589211940765, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0185, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.11124882102012634, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0141, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.10930055379867554, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0148, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.10555030405521393, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0152, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.0836755633354187, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0153, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.11849679052829742, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0158, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.13186359405517578, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.0167, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.11089541018009186, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0189, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.09473541378974915, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0155, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.1151607483625412, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0138, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.10416971892118454, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0138, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.10269354283809662, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0158, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11249537020921707, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0152, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.14352446794509888, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0169, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10745783895254135, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0159, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.09522922337055206, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0164, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.13701501488685608, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0146, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.08020906895399094, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0139, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09878123551607132, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.015, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.09399155527353287, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.015, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.11354538053274155, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0175, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.09135859459638596, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0156, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.08270548284053802, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.014, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.11252828687429428, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0157, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.0985439196228981, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0163, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.12125496566295624, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0146, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.09888023883104324, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0176, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.10250411182641983, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.10701503604650497, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0151, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11525820940732956, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0157, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.09755205363035202, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0159, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09472718834877014, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.015, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10360235720872879, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.0149, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.09946251660585403, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0164, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.12811636924743652, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0135, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.127360999584198, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.016, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.12480203062295914, + "learning_rate": 1.45003312024933e-05, + "loss": 0.0142, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.10674259066581726, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0156, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.12259232252836227, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0149, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.08998828381299973, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0149, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.10316051542758942, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0145, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.14545974135398865, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0158, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.1868322491645813, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.015, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.12345045059919357, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0154, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.11794525384902954, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0177, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.12257082760334015, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0152, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.12155821174383163, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0159, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12235453724861145, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0147, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.1056676059961319, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.015, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.1309097409248352, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0145, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.11743862181901932, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0152, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.12476198375225067, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.016, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.08397012203931808, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0162, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12207506597042084, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0159, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.09591932594776154, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0137, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.12573577463626862, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0164, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1315455585718155, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0164, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.1468929648399353, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0165, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.15008211135864258, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0148, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.09235277771949768, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0153, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.1275569051504135, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0163, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.08151543140411377, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0129, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10390543937683105, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.015, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.11043227463960648, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0141, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.1271016001701355, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0163, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09329107403755188, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0134, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.10737643390893936, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0135, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12903659045696259, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0159, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.11276338994503021, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0147, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.14026683568954468, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0163, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.12503503262996674, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0172, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.10071003437042236, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0138, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10627497732639313, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0138, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.0770413801074028, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0135, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.09400150179862976, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0154, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.1549134999513626, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.015, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.09969423711299896, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.0135, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.09770094603300095, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0149, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.0848713293671608, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0128, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08489260822534561, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0142, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.14606203138828278, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.0144, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.1283080279827118, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0147, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.09238846600055695, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0154, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.09167003631591797, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0153, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.1081148013472557, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0142, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.19600129127502441, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0158, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.08589469641447067, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0148, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10786846280097961, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0123, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.12004593759775162, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0154, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10496815294027328, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0133, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.1324186772108078, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0152, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.10160262882709503, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0137, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.14738032221794128, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0167, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.1030116155743599, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0146, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.10793277621269226, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0156, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.11900396645069122, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0165, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.12115667760372162, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0144, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.12846702337265015, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0163, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.09553872793912888, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.015, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.12807151675224304, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0161, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.12759897112846375, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.017, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.10901748389005661, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0151, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.09706093370914459, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0129, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10346078872680664, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.0147, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.12170220911502838, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0136, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.24494236707687378, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0139, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.10071292519569397, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0131, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.13647668063640594, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0151, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09296777099370956, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.0147, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.1015978679060936, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.0147, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.09014145284891129, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0134, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.08798721432685852, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0136, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.08899430930614471, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.014, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.10661277174949646, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.0189, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.10300924628973007, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.0165, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.10107792913913727, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0152, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.12540587782859802, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0147, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.14788442850112915, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0154, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.1628517359495163, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0155, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.12338890135288239, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0156, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09656467288732529, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0132, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.15762092173099518, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0142, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.1834212690591812, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0156, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.091162770986557, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.0142, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.14656929671764374, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0163, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.08447254449129105, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0144, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10864629596471786, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0137, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.12742409110069275, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0134, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.09852556139230728, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.13655197620391846, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0143, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.09299396723508835, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0134, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.08303989470005035, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0136, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.09860356897115707, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.016, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.08682689070701599, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0134, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.09411480277776718, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.0149, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.10258188843727112, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.016, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.09782259166240692, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0165, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.10300199687480927, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0148, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.09208223968744278, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0175, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.09144087135791779, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0148, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.09090441465377808, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0153, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.10571702569723129, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0168, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.1070566177368164, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0154, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.12103649973869324, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.015, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.10678276419639587, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0165, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11873649060726166, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0154, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.11791808903217316, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0145, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.0986858382821083, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.016, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09249810129404068, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0127, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12782645225524902, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0154, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.10502453148365021, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0162, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.09005366265773773, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.0143, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11655479669570923, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0158, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.10148292779922485, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0155, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.11018893122673035, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0147, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.11230522394180298, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0148, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.09911413490772247, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0159, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.09358150511980057, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0154, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.12689724564552307, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0145, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.10737589001655579, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.013, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.12670603394508362, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0148, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.1459159404039383, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0162, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.14686602354049683, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0139, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10709210485219955, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0161, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.08311928808689117, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.015, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12497760355472565, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0161, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.08203094452619553, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0129, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.12578807771205902, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0145, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10697296261787415, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0147, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.10980327427387238, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0154, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.09966560453176498, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0126, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.09740720689296722, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0126, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.1452278196811676, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.013, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.10068526864051819, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0152, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.10273153334856033, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0134, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.1101747378706932, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0148, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.11141166090965271, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0133, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1535777896642685, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0155, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.08465548604726791, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0161, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.07614064961671829, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0149, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09852079302072525, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0135, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.13796597719192505, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0163, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1669035404920578, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0143, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09881193935871124, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0142, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.12325254082679749, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0132, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.16859179735183716, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0168, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.09475059062242508, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.014, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.11981360614299774, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0186, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.10775262862443924, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0151, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.10053553432226181, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0129, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.10618098825216293, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0154, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.12225139886140823, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0171, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.07435912638902664, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0152, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.10211736708879471, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0177, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12564173340797424, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.015, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.10412495583295822, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0157, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.12016817182302475, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0146, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.10957430303096771, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0141, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08035340160131454, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0146, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.1221131756901741, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0153, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.10435835272073746, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0133, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.11107036471366882, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0137, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.0575222373008728, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0149, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.076436348259449, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0159, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.09325376152992249, + "learning_rate": 1.380704261284659e-05, + "loss": 0.0163, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.11715361475944519, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0161, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.08567722141742706, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0149, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.12162303924560547, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.017, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.11919409781694412, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.0154, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.10694293677806854, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0138, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.10736183077096939, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0133, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11603212356567383, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0127, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.1180153489112854, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.0161, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.11968789249658585, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.0151, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.1250530481338501, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0136, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10276933014392853, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0174, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.14728401601314545, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.0138, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.09268351644277573, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0129, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12262973934412003, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0147, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.11068148910999298, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0157, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.09095525741577148, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0149, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.12430994212627411, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0154, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1246076300740242, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0153, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.08052714169025421, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.0144, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.1667853593826294, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0164, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.10500273108482361, + "learning_rate": 1.372053087451907e-05, + "loss": 0.013, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.10832768678665161, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0171, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.11245214939117432, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0158, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.14754413068294525, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0146, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.16368411481380463, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0175, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.1533215045928955, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0138, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.11878912895917892, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0134, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.10748414695262909, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0155, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.14132313430309296, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0139, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.09151305258274078, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0133, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10014014691114426, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0143, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.12393905967473984, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0154, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13815879821777344, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0155, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.11488667875528336, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.014, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.08504864573478699, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0149, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.12521564960479736, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0161, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.09525968134403229, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0137, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.12103450298309326, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.017, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.10642504692077637, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0139, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.15175847709178925, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0153, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11594653874635696, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.017, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.12788720428943634, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0171, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.12372581660747528, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0157, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.09206387400627136, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0155, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.09841199964284897, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0162, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.135525643825531, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0148, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.0982549637556076, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.0148, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.07432328909635544, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0124, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09830821305513382, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0139, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.10874123871326447, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0149, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.08979356288909912, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0189, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.08131087571382523, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.0128, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.1204582154750824, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0141, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11884505301713943, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0144, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.16451388597488403, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0134, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.11953870207071304, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0135, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12111926078796387, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0159, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.1181047335267067, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0142, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.12142453342676163, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0156, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09890097379684448, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.0137, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.13294243812561035, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0132, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.1397709846496582, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0161, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.10867296159267426, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.014, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11157155781984329, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.015, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09516428411006927, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0156, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.21431738138198853, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0158, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.09854292124509811, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0159, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.21810708940029144, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0152, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.14076393842697144, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0162, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.0892275720834732, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0154, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.09736334532499313, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0127, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.13880307972431183, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.015, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1091906875371933, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0147, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11605603992938995, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0168, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.11140985041856766, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0162, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.09057680517435074, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0161, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.12626419961452484, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0161, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.136195570230484, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0142, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.12415066361427307, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0121, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.12581513822078705, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.0151, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.11835133284330368, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0147, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.12255530059337616, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.0135, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10576335340738297, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0149, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.12575241923332214, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.014, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09812266379594803, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0152, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.11845924705266953, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0143, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.10975389182567596, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0147, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.09649760276079178, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0131, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.12912778556346893, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.0152, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10225717723369598, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0143, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.1712195724248886, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0144, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.1477300226688385, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0155, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.11692417412996292, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0146, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.1232081949710846, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0152, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.14507624506950378, + "learning_rate": 1.341604094324139e-05, + "loss": 0.017, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.13909533619880676, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0136, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.1438206136226654, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.0167, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.09637414664030075, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0134, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.15815401077270508, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.015, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.12961018085479736, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0159, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.11171260476112366, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0133, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.0907273143529892, + "learning_rate": 1.33872699666556e-05, + "loss": 0.015, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.11181363463401794, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0162, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.11502614617347717, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0169, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.14428892731666565, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0158, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.1440858244895935, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0134, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.07517627626657486, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0129, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.11884543299674988, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0147, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.08753574639558792, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0178, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.09831796586513519, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.014, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.12262003868818283, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.016, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.11444123834371567, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0142, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11559469252824783, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0138, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.09936029464006424, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0116, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.12810789048671722, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.0137, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.07923603802919388, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0136, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.09709552675485611, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0148, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.0924353376030922, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0152, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.10444460809230804, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0136, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.09510108828544617, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.014, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.11014958471059799, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.014, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.13198788464069366, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0153, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.09857133030891418, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0128, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.15041643381118774, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0148, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.13857626914978027, + "learning_rate": 1.329277958632556e-05, + "loss": 0.018, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.13004948198795319, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0142, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.09864714741706848, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0143, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12682363390922546, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0177, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.09303848445415497, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0141, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.13559986650943756, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0153, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.10143767297267914, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0131, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.1287422627210617, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0142, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.12000516802072525, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.016, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.10434254258871078, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.014, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08141680061817169, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0138, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08781418204307556, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0138, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.1355503350496292, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.015, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.1243978887796402, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0143, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.11088231950998306, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0154, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12559033930301666, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0132, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.16149592399597168, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.09809687733650208, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.0142, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.12586510181427002, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0165, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.12403467297554016, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.14458157122135162, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.0135, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.09313284605741501, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0165, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.14337806403636932, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0144, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.12650802731513977, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0126, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.112495556473732, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.0154, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11324562877416611, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.0139, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.11720716208219528, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.0146, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.1261121928691864, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0143, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.1302812099456787, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0157, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.12074442952871323, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0172, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.11716746538877487, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0158, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.11565797030925751, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0144, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.11653099209070206, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0155, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.13103044033050537, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0152, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.10341408103704453, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0138, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.11098334193229675, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.0146, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.11139992624521255, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.10786942392587662, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0143, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.10139934718608856, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0154, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.12619639933109283, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0153, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.13168810307979584, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0154, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.09545969218015671, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0151, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.09984135627746582, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0129, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10903751850128174, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0135, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.11882342398166656, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0151, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.1105988621711731, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0137, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.09874245524406433, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0142, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.10673530399799347, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0157, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.10264459252357483, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0162, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10769633203744888, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0133, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.16478870809078217, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0149, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.12630558013916016, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0128, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09824837744235992, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0144, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12136032432317734, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0124, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.10943865031003952, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0154, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.11394404619932175, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.0117, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.10907739400863647, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0144, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10671219974756241, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0156, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.10080806165933609, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0141, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.17395640909671783, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0145, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.12620997428894043, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0186, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.13434381783008575, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0149, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.09921885281801224, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.0149, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.13742485642433167, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0155, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.0998004674911499, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10389069467782974, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0138, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.11996421217918396, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.0147, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.09136862307786942, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0129, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.11111150681972504, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0159, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.11024610698223114, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0138, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.15376223623752594, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0154, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.13934020698070526, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0138, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.11208872497081757, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0131, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1513150930404663, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.10351566225290298, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0137, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11078765988349915, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0164, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.15622863173484802, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.0147, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.11603175848722458, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0145, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.1254235804080963, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0162, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.0900387316942215, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.015, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09852428734302521, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0172, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.13502047955989838, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.014, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.1460445076227188, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0156, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.10317493230104446, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0129, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.127616748213768, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0157, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.11560963094234467, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0142, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.11239839345216751, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0135, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.1503314971923828, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0157, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.1253356635570526, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0154, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.13217659294605255, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0127, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09087327867746353, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0144, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.09290582686662674, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0147, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.10273045301437378, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0165, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.14741986989974976, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.016, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.0804770216345787, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0125, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.12413053959608078, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0136, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.09822124987840652, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0176, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.09091541171073914, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0161, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.09487786144018173, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0148, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1009742021560669, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0147, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.08731451630592346, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0118, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.08123810589313507, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0134, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.10559827089309692, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0159, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.11313227564096451, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0164, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.11703694611787796, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0171, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.1083052009344101, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0168, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.13762544095516205, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0185, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.07979343086481094, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.09789150208234787, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0139, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.1905684620141983, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0158, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.1036764457821846, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0118, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.09413912892341614, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0137, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.2435447722673416, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.0132, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.09290791302919388, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0146, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.1211535856127739, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0151, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.1133081242442131, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.013, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.11362706869840622, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0171, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.10019481182098389, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0151, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.09489522129297256, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0143, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12912064790725708, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0138, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.11348122358322144, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0136, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.1325748860836029, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0156, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.11978204548358917, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0146, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.10022053122520447, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0137, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.07388417422771454, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0152, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.10844552516937256, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.0142, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.12381294369697571, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0132, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09625408798456192, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.0154, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.1207597553730011, + "learning_rate": 1.276827841111951e-05, + "loss": 0.016, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.12216818332672119, + "learning_rate": 1.276419082873553e-05, + "loss": 0.016, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.0969374030828476, + "learning_rate": 1.27601034161465e-05, + "loss": 0.0125, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.13043035566806793, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0145, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.10600738227367401, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0144, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.11113949120044708, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0135, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.09458228200674057, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0139, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10076881945133209, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0139, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1297597885131836, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0142, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.10059160739183426, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0142, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10716533660888672, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0149, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.1138077974319458, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0151, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.13693730533123016, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0136, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.08157949149608612, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0143, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.09702442586421967, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0121, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09029403328895569, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0133, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.1444508135318756, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0133, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.11026642471551895, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0161, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.11229082942008972, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0135, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.08556455373764038, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0166, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10684676468372345, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.0142, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.11970192193984985, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0132, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.10227229446172714, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.015, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.08301195502281189, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0153, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.1275268793106079, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.014, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.11613821983337402, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.015, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.10491939634084702, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0127, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.09623268246650696, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0136, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.111306332051754, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0137, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0863005742430687, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.0137, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.09434706717729568, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0142, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.10771634429693222, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0155, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.14157304167747498, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.0145, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11419932544231415, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0154, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.12308389693498611, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0133, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.10775505006313324, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.0144, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.10383253544569016, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0141, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.1012682244181633, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0144, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.12071908265352249, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0157, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.12535089254379272, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0149, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.08634085953235626, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0152, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.10035993158817291, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0138, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.13245077431201935, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0135, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.08713636547327042, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0144, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.10049428045749664, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0134, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.10432936251163483, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0163, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.16578327119350433, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0166, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.16106748580932617, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0125, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.11968962103128433, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0142, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1189485490322113, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0155, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.10795646905899048, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0139, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.08580315858125687, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0134, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.17969708144664764, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0159, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09675301611423492, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0146, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.10921840369701385, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0158, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.11278913170099258, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0133, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12176436185836792, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0138, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.1186685562133789, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0171, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.10296562314033508, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.015, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.15394653379917145, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0173, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.06815146654844284, + "learning_rate": 1.252333462101563e-05, + "loss": 0.014, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.1100328117609024, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0129, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.08578243851661682, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0145, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.11262256652116776, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.015, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.09644307941198349, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0183, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09589987248182297, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0145, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.09658264368772507, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0144, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.11963947862386703, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0118, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.14156748354434967, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0164, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.14387111365795135, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0151, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.11542942374944687, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0123, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.12162452191114426, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0165, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.13585466146469116, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0157, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.11272095143795013, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0159, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.1828201562166214, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.0162, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.09920861572027206, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0163, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.12618015706539154, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.016, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.13111864030361176, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0155, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.08486807346343994, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0126, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.14975349605083466, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0156, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.12495499104261398, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0154, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.11145244538784027, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0147, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.08242745697498322, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0149, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.10702498257160187, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0152, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.16176027059555054, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0145, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.0959012433886528, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0134, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10282868891954422, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0143, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.11747556924819946, + "learning_rate": 1.241332681912234e-05, + "loss": 0.0165, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10770756006240845, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0134, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.09925016760826111, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0126, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.09207499027252197, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.016, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.21546097099781036, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0131, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.10848663747310638, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0182, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.09382427483797073, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0134, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.10758832842111588, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0136, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.10075095295906067, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0154, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.09366318583488464, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0151, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.13190370798110962, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0154, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.1926490068435669, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0162, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.10018754750490189, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0127, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10250869393348694, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0146, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.08466067910194397, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0152, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.10397373139858246, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.0148, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14372816681861877, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0151, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11059669405221939, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0133, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09743087738752365, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0151, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.11265920102596283, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0145, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.08748266100883484, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0146, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.10545652359724045, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0133, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.09680107235908508, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0149, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.13201938569545746, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0152, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.12063595652580261, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0167, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.09986719489097595, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0153, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.09913598746061325, + "learning_rate": 1.230752865632506e-05, + "loss": 0.0151, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.1373804211616516, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0144, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.12088140845298767, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0137, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.08200784027576447, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0145, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.12470699846744537, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0154, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.10114862769842148, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.13169868290424347, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0159, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.08321736752986908, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0147, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.10108216106891632, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.012, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.08475375175476074, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0155, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.08946104347705841, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0145, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.07863342016935349, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0143, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.0906776711344719, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0123, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.11063490062952042, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0128, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.10001818835735321, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0133, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.12135739624500275, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0145, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.1233866810798645, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0133, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.09280609339475632, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0161, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.11677120625972748, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0144, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.11756281554698944, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.0163, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.08084359019994736, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.1415908932685852, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0149, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09107301384210587, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.013, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.11712655425071716, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.014, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.08366917073726654, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0125, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10391885787248611, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0146, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.1042654886841774, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0132, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.09504523873329163, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0148, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.12793844938278198, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0164, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.16210904717445374, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.15506179630756378, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.0153, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.11274553835391998, + "learning_rate": 1.218156580082751e-05, + "loss": 0.015, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.14852653443813324, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0172, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13558489084243774, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0156, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.08902081102132797, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0152, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12315832078456879, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0154, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.13528762757778168, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0145, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.1538013368844986, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0151, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.09894628077745438, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0154, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07696802169084549, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0136, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.09229687601327896, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0131, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.11709824949502945, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0146, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09195996820926666, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0142, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.11714771389961243, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.0157, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.11561911553144455, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0162, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.12904389202594757, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0129, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.10347782075405121, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0126, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.10054155439138412, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0126, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.1713176816701889, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0142, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10279975831508636, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0133, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.11513496935367584, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0137, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.096926249563694, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0123, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.1159687489271164, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.0145, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09745503962039948, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.014, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.15608133375644684, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.0144, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.08617804199457169, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.09285406023263931, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0123, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.11964505165815353, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0143, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1376206874847412, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0136, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.12900781631469727, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0148, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.10527875274419785, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0157, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.11285959929227829, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0135, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.09756818413734436, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0158, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.10695865005254745, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09360404312610626, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0129, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.100432388484478, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0126, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.10982918739318848, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0158, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.07767278701066971, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.015, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.10380839556455612, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0153, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.1225915253162384, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0167, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.0981304869055748, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0139, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.10055239498615265, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0146, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.09311894327402115, + "learning_rate": 1.201528720560666e-05, + "loss": 0.016, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.11859868466854095, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0163, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.15336444973945618, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0153, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.09175246208906174, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0125, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.0973501205444336, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0155, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.1643284410238266, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0142, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.1077687218785286, + "learning_rate": 1.19909852066355e-05, + "loss": 0.015, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.07722865790128708, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0125, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.149483323097229, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.015, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.11405838280916214, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0144, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.11132190376520157, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0142, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.10609457641839981, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0128, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.09458910673856735, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0134, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.10288971662521362, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0164, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.0970534011721611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0128, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.09571509063243866, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0129, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.08922260254621506, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0144, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0886709988117218, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.017, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.11732205003499985, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0141, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.11602790653705597, + "learning_rate": 1.193835918571248e-05, + "loss": 0.016, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.11643379926681519, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.0156, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.0927688404917717, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0149, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11533524841070175, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0133, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11001960933208466, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0135, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.12629783153533936, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.014, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.13336558640003204, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0121, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.11672402173280716, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.016, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.10014890134334564, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0136, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.0940401554107666, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.0151, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.1272585242986679, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0139, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.08492123335599899, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0151, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.1195707842707634, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0137, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.1095622330904007, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0167, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.09656022489070892, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0137, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1358073204755783, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0166, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.10981698334217072, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0141, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.07301078736782074, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0127, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.1040695309638977, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0143, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.16375352442264557, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.0136, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.19000792503356934, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0172, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11925465613603592, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0148, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.188803032040596, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.0151, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.09543585777282715, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0129, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11177625507116318, + "learning_rate": 1.184130718987356e-05, + "loss": 0.013, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.11703382432460785, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0153, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.12326977401971817, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0151, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.09018218517303467, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0147, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1641266793012619, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0146, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.08780542016029358, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0127, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09138940274715424, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0133, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.11768870800733566, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0128, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10298353433609009, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0141, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.10553838312625885, + "learning_rate": 1.180494808593063e-05, + "loss": 0.014, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10878341645002365, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0166, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10571706295013428, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0166, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.09349539875984192, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0126, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.10237022489309311, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0128, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.11938492953777313, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0167, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.10417849570512772, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0132, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.10322093963623047, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0141, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.09445950388908386, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.0146, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.1363389790058136, + "learning_rate": 1.176860863606944e-05, + "loss": 0.014, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.10975046455860138, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.013, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.12059775739908218, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0136, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.10529737174510956, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0158, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.11280884593725204, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0146, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.09665497392416, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0126, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.14059090614318848, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0142, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.10380115360021591, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.0132, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.10641331225633621, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0157, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.10845568776130676, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0135, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.10574200004339218, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0153, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.09128011763095856, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0142, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.10704760998487473, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0129, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.1180003359913826, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0141, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.11784005165100098, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0138, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.08203493803739548, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0144, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.10125461220741272, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.0142, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.12113944441080093, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0124, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.09266798943281174, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0156, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.11875846236944199, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0146, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.11041444540023804, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.013, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.13168254494667053, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0151, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.08208412677049637, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0132, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.08591622114181519, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0152, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.14605054259300232, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.014, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.12816093862056732, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0163, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.10751980543136597, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.013, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.1012972742319107, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0134, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13583813607692719, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0121, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.13573803007602692, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0143, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11719135195016861, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.0163, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.11966057121753693, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0133, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.11172359436750412, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0146, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10357552021741867, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0134, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.09571186453104019, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0143, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.13586629927158356, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0162, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.10461284965276718, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0134, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.182363823056221, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.014, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.12570780515670776, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0161, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.10319641977548599, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0136, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1066908910870552, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0123, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.10409434139728546, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0143, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.08230862766504288, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0128, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.110751211643219, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0148, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.09656667709350586, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0137, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.10626243054866791, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0127, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10294152796268463, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0151, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.09571471065282822, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0136, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.1162625253200531, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0129, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.09755981713533401, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0151, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.08586591482162476, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.016, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.11462593078613281, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.013, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.11438485980033875, + "learning_rate": 1.155904354214324e-05, + "loss": 0.016, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.15895701944828033, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0161, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.1135995015501976, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0128, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.11993709206581116, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0142, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.14105743169784546, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0118, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.08521834015846252, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0136, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.17094957828521729, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0159, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.1379796713590622, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.012, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.10417291522026062, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0138, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.09216596931219101, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0151, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.08398481458425522, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0124, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.1291581243276596, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0152, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.09349840134382248, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0127, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.08282706141471863, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0143, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.10378136485815048, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0146, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1504564881324768, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0144, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.0940636470913887, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0147, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.11142407357692719, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0127, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.16776388883590698, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0158, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.10488522052764893, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0132, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.09612040966749191, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0119, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.09318249672651291, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0116, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.08647014945745468, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0124, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.1067274957895279, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0136, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.08162420988082886, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0136, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.08949539810419083, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0136, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.11327249556779861, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0158, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.10963857173919678, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0141, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.10124490410089493, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.014, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10530558228492737, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0135, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.1251840591430664, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0127, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.14418582618236542, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0156, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.09544084221124649, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0137, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.09794268012046814, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0151, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.10741079598665237, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0127, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.10632490366697311, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0117, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08640208840370178, + "learning_rate": 1.141437135633602e-05, + "loss": 0.0119, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.09476958960294724, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0134, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.1141064390540123, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0147, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.1008499413728714, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0119, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.0853564441204071, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.12970876693725586, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0128, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11144055426120758, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0137, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.09443596750497818, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0127, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.19357536733150482, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0173, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.11861733347177505, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0145, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.13152548670768738, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0145, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11607249826192856, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0128, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.09456845372915268, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0145, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.12110379338264465, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.014, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.07872474938631058, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0128, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.1042177751660347, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0136, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.15168064832687378, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0126, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.09023087471723557, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0125, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.1110723614692688, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0121, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.09889500588178635, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0136, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.13798406720161438, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0151, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.14507459104061127, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0175, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10034530609846115, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0143, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.09978580474853516, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0142, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.11770957708358765, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0145, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.10433546453714371, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0148, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.1006259173154831, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.016, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.09567620605230331, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0133, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08685444295406342, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0116, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.10901349037885666, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.0137, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.15114818513393402, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0147, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.08353722840547562, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0146, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.14387859404087067, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0135, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.12125595659017563, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0148, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.11398079991340637, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0142, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.14072677493095398, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0149, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.11925073713064194, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0153, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.10301142185926437, + "learning_rate": 1.126604838049735e-05, + "loss": 0.013, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.11580533534288406, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.012, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.10776577144861221, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0143, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.1724127233028412, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0146, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.15091492235660553, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0128, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.10073890537023544, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0131, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11136317253112793, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0127, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.11012828350067139, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0133, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.12291619181632996, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0148, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.15791606903076172, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0144, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.12378063052892685, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.0137, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12597671151161194, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0172, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.1783512681722641, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0169, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.07375630736351013, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0139, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.07327539473772049, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0125, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.09731314331293106, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0144, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.0931868702173233, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0124, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.10977564752101898, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0132, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.0881611704826355, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0168, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.1239505335688591, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0133, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.10325739532709122, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0132, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11499917507171631, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0119, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.10131840407848358, + "learning_rate": 1.117803938925701e-05, + "loss": 0.014, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.10435229539871216, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0119, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.10805722326040268, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0136, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0759270042181015, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0167, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.1406354457139969, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0155, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1452491581439972, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.10772743821144104, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0148, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09850231558084488, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0155, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.08501581847667694, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0155, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11223053187131882, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0146, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.07870364189147949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0136, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.11768847703933716, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.0153, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.096342071890831, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0137, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.10153007507324219, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0139, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.0837850570678711, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0117, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.11286687850952148, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0148, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.14671675860881805, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0117, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.09201597422361374, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0161, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.11397483944892883, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0127, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.08533219248056412, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0126, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.18251192569732666, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0134, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.17834120988845825, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.014, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11550542712211609, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0127, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11998283863067627, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0149, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11872667074203491, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0155, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.10344526916742325, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.015, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.11288250237703323, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0149, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.16417740285396576, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0128, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09344452619552612, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0131, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11792241781949997, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0159, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10204951465129852, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0143, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.13064894080162048, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0151, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.08712288737297058, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0148, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.11878253519535065, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0123, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0907600000500679, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0135, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.08273347467184067, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0143, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.08222810924053192, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0112, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.11497339606285095, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0166, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.115402452647686, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0123, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.09611446410417557, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0128, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.12478714436292648, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0133, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.1713663637638092, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0118, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.13887059688568115, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0144, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.08571378141641617, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0154, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.1155383437871933, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0172, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.1123170331120491, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0153, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.13517709076404572, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0142, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.10638624429702759, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0121, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.11200621724128723, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0129, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10821404308080673, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.0128, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.10840310156345367, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0119, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.10352504998445511, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0144, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.10084252059459686, + "learning_rate": 1.097058241668254e-05, + "loss": 0.0128, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.11881101131439209, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.09931256622076035, + "learning_rate": 1.09626195543497e-05, + "loss": 0.0148, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.1260150521993637, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0125, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14202995598316193, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0133, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.15601785480976105, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0158, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.10701733827590942, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0139, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10180657356977463, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0138, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.0962422713637352, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.0124, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.13017651438713074, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0153, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.1295875608921051, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0154, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.10589175671339035, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0144, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.10173171758651733, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0142, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11211177706718445, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.0144, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.09875764697790146, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.014, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.11233554035425186, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0135, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.11327439546585083, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0136, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.10059859603643417, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0143, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.1160704717040062, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0132, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.15698356926441193, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0163, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.08235745877027512, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0134, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.12417344003915787, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0145, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.25982916355133057, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0155, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.12248262017965317, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0139, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.11425210535526276, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.015, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.12354899197816849, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0187, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.1102118268609047, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0137, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11539250612258911, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.015, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.12405017763376236, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0136, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.10405998677015305, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.014, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.0918804481625557, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0162, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.10429271310567856, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0128, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.09917982667684555, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0133, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.15390615165233612, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0133, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.11222631484270096, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0151, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.14449205994606018, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0165, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1332763433456421, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.015, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09963493794202805, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0132, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.12338289618492126, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0122, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.09369654208421707, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0127, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.13490226864814758, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0149, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.08246663957834244, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0158, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.0892438217997551, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0124, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.1275915652513504, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0135, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.150055930018425, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0118, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.08317377418279648, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0141, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.118314728140831, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.0131, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.086687833070755, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0128, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.10310143977403641, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0125, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.06559336930513382, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0135, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.11680027842521667, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0135, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.1381155252456665, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.015, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.08212172240018845, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0139, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.11066312342882156, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0145, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.11565965414047241, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0131, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.08814681321382523, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0122, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.11116618663072586, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0156, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.10491067916154861, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0157, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.15800951421260834, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0141, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.10169819742441177, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0156, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.07719215005636215, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0127, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.08958051353693008, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0145, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.1171223521232605, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0141, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.09662648290395737, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0121, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.09235699474811554, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0129, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.127174973487854, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0157, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.10298880189657211, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0146, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08706901967525482, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0132, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.10898775607347488, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0117, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.12178633362054825, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0129, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.10881572961807251, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0138, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.12326113879680634, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.013, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.13331058621406555, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0129, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.11921956390142441, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0154, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.10736659914255142, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.014, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09894116222858429, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0141, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.09651704877614975, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0146, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.11836998164653778, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.0176, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.1108580157160759, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0129, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.1281266212463379, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0138, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.13581599295139313, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0149, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10911770164966583, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0162, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.14131440222263336, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0126, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.10660075396299362, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0144, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.12547564506530762, + "learning_rate": 1.063722589416218e-05, + "loss": 0.0142, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11465640366077423, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0143, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.1274002492427826, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0145, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.10782597959041595, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0144, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.08273816108703613, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0126, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.12750636041164398, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0113, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.10962793231010437, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0132, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11728726327419281, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0146, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.10677216202020645, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0146, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.0974075049161911, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.0129, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.09190504997968674, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0129, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.10864423215389252, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0122, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.08833710849285126, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0141, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.14306439459323883, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.013, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.11269631236791611, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0115, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.08850698918104172, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0116, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10025947540998459, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0116, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.07444360852241516, + "learning_rate": 1.057003955600053e-05, + "loss": 0.013, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.10606181621551514, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0133, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.11394638568162918, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0143, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.12015067040920258, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0128, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.11465302109718323, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0117, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.09164674580097198, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0116, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.10532126575708389, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0128, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.16767282783985138, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0137, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.16740484535694122, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0151, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08568812906742096, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0135, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.24568453431129456, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0138, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.13317807018756866, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0142, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11310465633869171, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0139, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.09677092730998993, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0123, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.13912886381149292, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0152, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.0902378037571907, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0124, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09430436044931412, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0116, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.13361187279224396, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0161, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09588316082954407, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0135, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.13104380667209625, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0161, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.11476746946573257, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0137, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.087586909532547, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0124, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.12136878073215485, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0156, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.09384764730930328, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0137, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.10422758758068085, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0142, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.11391496658325195, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0125, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.12965570390224457, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0137, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.13042569160461426, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.0159, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.10433127731084824, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0129, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.0989370346069336, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0132, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.08399394899606705, + "learning_rate": 1.045171352558859e-05, + "loss": 0.013, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.11693604290485382, + "learning_rate": 1.044777463361479e-05, + "loss": 0.0148, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.12304933369159698, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0121, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.10788783431053162, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0138, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11043704301118851, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0135, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.11343314498662949, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0158, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.10075241327285767, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0143, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.08752909302711487, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.012, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.11036614328622818, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0126, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.09241441637277603, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.0157, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.1369611620903015, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.0132, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.12782007455825806, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0136, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.140136256814003, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0144, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11508572846651077, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.0136, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.14334864914417267, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0138, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.10154326260089874, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0135, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.11290924996137619, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0134, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.13472969830036163, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0144, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.1215982660651207, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0125, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.10628669708967209, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0125, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.07950089126825333, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0135, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.14542941749095917, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0145, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.1189931258559227, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0115, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.11141669005155563, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0135, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.08776746690273285, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0136, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10942544043064117, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0139, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.15493649244308472, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0123, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.1485723853111267, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.0141, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.0886952131986618, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0124, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.09021062403917313, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0132, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.17778685688972473, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0126, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.11892827600240707, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0125, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.14066243171691895, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0135, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.10870271921157837, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0147, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.10636188089847565, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0156, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.11178954690694809, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.014, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.09059949964284897, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.0127, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.11947090923786163, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0145, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.1133369505405426, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.014, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.12451290339231491, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0139, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10219574719667435, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0138, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.09843198210000992, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0144, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.09784907102584839, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0155, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1259097307920456, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0156, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.08604080229997635, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0113, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.11517608165740967, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0131, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.09721313416957855, + "learning_rate": 1.027088684228894e-05, + "loss": 0.011, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.08404546976089478, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0164, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.11207510530948639, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0151, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.1101691946387291, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0129, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.2047414630651474, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0143, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.08887176960706711, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0139, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.08608511090278625, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0162, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10840177536010742, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0133, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.1228431761264801, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.0173, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.09171874076128006, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0119, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.1357734203338623, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0151, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.13374143838882446, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0134, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09826765209436417, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0112, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.30271241068840027, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0128, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.1474897861480713, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.0142, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.11385788768529892, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0121, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.1145087331533432, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0125, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.15277381241321564, + "learning_rate": 1.020425036775911e-05, + "loss": 0.011, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10591624677181244, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0123, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.09938671439886093, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0136, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.12457235157489777, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0145, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.10550948232412338, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0127, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.12241430580615997, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.013, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.12996822595596313, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0127, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.10886852443218231, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0124, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.10862760245800018, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0124, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.12071339040994644, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0152, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.12002567946910858, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0144, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.0731884092092514, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.014, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.09087152034044266, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0122, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.14147141575813293, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0126, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.10024842619895935, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.013, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.12605364620685577, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0129, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.07398458570241928, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.012, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.09070646017789841, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0139, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.06930165737867355, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.012, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.10134197026491165, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0131, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.12521374225616455, + "learning_rate": 1.012598938117799e-05, + "loss": 0.012, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.10238701850175858, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0139, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.10689418017864227, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0138, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.0954018235206604, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0125, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.1394306868314743, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0125, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.11217380315065384, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0143, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.1513020396232605, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.0122, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.11155983060598373, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0132, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.1257336288690567, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.015, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.10818599164485931, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0163, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.15947701036930084, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0142, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.10929335653781891, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0129, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.08994319289922714, + "learning_rate": 1.007910370568197e-05, + "loss": 0.011, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.09995368868112564, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0131, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.10167150944471359, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.0142, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1480625420808792, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0129, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.11566069722175598, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.013, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.10494371503591537, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0132, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.09531164914369583, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0116, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.0887216180562973, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0133, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.09848874062299728, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0122, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.14705902338027954, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0135, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.1412736475467682, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11742296814918518, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0143, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.11000296473503113, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0144, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.111237533390522, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0125, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.1085008904337883, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0133, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.12815521657466888, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0138, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.11128585785627365, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.12439922988414764, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0139, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.11863227933645248, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0154, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.13517265021800995, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0128, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.0960967019200325, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0123, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.0953422263264656, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0121, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.1965440809726715, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.0924379900097847, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0112, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.08345641940832138, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0146, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.10248678177595139, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0121, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.2980184257030487, + "learning_rate": 9.977703371038337e-06, + "loss": 0.013, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.11192528903484344, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0133, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.21329663693904877, + "learning_rate": 9.969913958481688e-06, + "loss": 0.0116, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.0922258049249649, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0159, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.12964527308940887, + "learning_rate": 9.962126073932371e-06, + "loss": 0.0143, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.11362354457378387, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0137, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.0902123972773552, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0126, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.08770131319761276, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0123, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.09604862332344055, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0131, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.08182749897241592, + "learning_rate": 9.942663063115248e-06, + "loss": 0.013, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.17937488853931427, + "learning_rate": 9.938771611983594e-06, + "loss": 0.0158, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.1343632936477661, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0139, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.14316903054714203, + "learning_rate": 9.93098986311647e-06, + "loss": 0.012, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11366818100214005, + "learning_rate": 9.927099565971976e-06, + "loss": 0.0129, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0949450209736824, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0144, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.13775621354579926, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0134, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.10221140086650848, + "learning_rate": 9.915430987239967e-06, + "loss": 0.012, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.11101876199245453, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0121, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.12785188853740692, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0146, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.1400655061006546, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0115, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.09468253701925278, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0122, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.08724251389503479, + "learning_rate": 9.895991082746168e-06, + "loss": 0.0143, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.10251675546169281, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.11980220675468445, + "learning_rate": 9.888217832192542e-06, + "loss": 0.015, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.12473925948143005, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0144, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.08260339498519897, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0122, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.11015481501817703, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0154, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1120007336139679, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0143, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09239251911640167, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0146, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10413528233766556, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0144, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.0863887220621109, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0143, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.0963623896241188, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0129, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.2090010792016983, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0155, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12460873275995255, + "learning_rate": 9.84937491900087e-06, + "loss": 0.012, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.11049462109804153, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0134, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.15180432796478271, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0142, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.13392038643360138, + "learning_rate": 9.837729659166156e-06, + "loss": 0.013, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.11534144729375839, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0134, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.1077096238732338, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0135, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.10903865098953247, + "learning_rate": 9.826087927694543e-06, + "loss": 0.014, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.09442178905010223, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0132, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.12171642482280731, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0143, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.08623645454645157, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0135, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.10249772667884827, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0151, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.16136185824871063, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0113, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.15169787406921387, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0161, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.095759317278862, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0143, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.11120729893445969, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0127, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09804334491491318, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0123, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.11462689191102982, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0111, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10223866254091263, + "learning_rate": 9.783431894635035e-06, + "loss": 0.012, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.1001039668917656, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0149, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.13274405896663666, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0136, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.09141503274440765, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0135, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.1291927844285965, + "learning_rate": 9.767932474193758e-06, + "loss": 0.0136, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10977526009082794, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0135, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.1167440190911293, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0138, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.08445189893245697, + "learning_rate": 9.756312079913248e-06, + "loss": 0.0129, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.2519867718219757, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0146, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.08974999934434891, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0152, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.1252872496843338, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0147, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.16116158664226532, + "learning_rate": 9.740823797373594e-06, + "loss": 0.013, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.08403292298316956, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0119, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.11238662898540497, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0154, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10106278210878372, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0136, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.08145973831415176, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0131, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.15683840215206146, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0132, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.14468176662921906, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0146, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.08105593919754028, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0122, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.12362894415855408, + "learning_rate": 9.709866422291219e-06, + "loss": 0.013, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.09933894127607346, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0156, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.09758161753416061, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0145, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.1307433545589447, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0139, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.1336813122034073, + "learning_rate": 9.694397367364635e-06, + "loss": 0.015, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.08841264992952347, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0162, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09142819046974182, + "learning_rate": 9.68666525627968e-06, + "loss": 0.0132, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11312894523143768, + "learning_rate": 9.682799805859474e-06, + "loss": 0.0129, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.14984415471553802, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0147, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.104182668030262, + "learning_rate": 9.675070116731382e-06, + "loss": 0.013, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.10078790783882141, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0131, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.08538635075092316, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0135, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.11059115827083588, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0134, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.10521579533815384, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0131, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.1130848377943039, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0137, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.09023619443178177, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0125, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.11787812411785126, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0136, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.14386765658855438, + "learning_rate": 9.6441675594234e-06, + "loss": 0.0153, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.10471580922603607, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0128, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.1340251863002777, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0139, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.1131536215543747, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0128, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10629499703645706, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0147, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.10359197854995728, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0121, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.11530483514070511, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0136, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.09336108714342117, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0124, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.08179444819688797, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0124, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.12341088801622391, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0125, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.12123409658670425, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0149, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.11406739801168442, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0141, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.07922708988189697, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0114, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.09156256914138794, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0121, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.09705155342817307, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0134, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.10157317668199539, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0157, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.09271441400051117, + "learning_rate": 9.582440688952473e-06, + "loss": 0.0113, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.09375326335430145, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0109, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.07997307926416397, + "learning_rate": 9.57473221124502e-06, + "loss": 0.0137, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.11743918806314468, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0138, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.09150344878435135, + "learning_rate": 9.567025381590669e-06, + "loss": 0.013, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.09968695789575577, + "learning_rate": 9.563172585515032e-06, + "loss": 0.011, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.10516709089279175, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0114, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.09716377407312393, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0132, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.12646295130252838, + "learning_rate": 9.551616675805319e-06, + "loss": 0.0145, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.11036434769630432, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0153, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10279295593500137, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0138, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.08990927040576935, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0132, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.10534624755382538, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0138, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.09507470577955246, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0114, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.101425401866436, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0137, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.11853020638227463, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0121, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.13193756341934204, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0128, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.10581472516059875, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0123, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.12424535304307938, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0139, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.1036083921790123, + "learning_rate": 9.509276927101462e-06, + "loss": 0.0126, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.07990803569555283, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0112, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.098060242831707, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0136, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.11892619729042053, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0126, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.10609585791826248, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0127, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.10905738174915314, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0123, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.0981525182723999, + "learning_rate": 9.486203760067013e-06, + "loss": 0.012, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.16955061256885529, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0119, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.10360468924045563, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0136, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.08537023514509201, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0112, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.1317787766456604, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0128, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10730036348104477, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0151, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09596534818410873, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0117, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.10839905589818954, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0138, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.12511657178401947, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0138, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.1078735962510109, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0127, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.10344097018241882, + "learning_rate": 9.447782011716307e-06, + "loss": 0.0119, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.18708902597427368, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0139, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.09599670022726059, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0125, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.13859155774116516, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0143, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09208878129720688, + "learning_rate": 9.432425102360072e-06, + "loss": 0.0126, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.21019718050956726, + "learning_rate": 9.428586931343613e-06, + "loss": 0.0117, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.09587983787059784, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0122, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.09626271575689316, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0124, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.1051478385925293, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0154, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.10166551172733307, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0119, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.09312118589878082, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0145, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.11722584068775177, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0126, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.0939759686589241, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0124, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.11434069275856018, + "learning_rate": 9.39789681972375e-06, + "loss": 0.012, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12582115828990936, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0137, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.07123387604951859, + "learning_rate": 9.390228540420876e-06, + "loss": 0.0155, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.08926647901535034, + "learning_rate": 9.386395039661539e-06, + "loss": 0.014, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.07371679693460464, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0128, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.1008603498339653, + "learning_rate": 9.378729317382715e-06, + "loss": 0.012, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.10988970100879669, + "learning_rate": 9.37489709644538e-06, + "loss": 0.0136, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.09886831045150757, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0123, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.10959886759519577, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0135, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.12443319708108902, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0161, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10207810997962952, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0146, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.11660879850387573, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0127, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.13432076573371887, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0142, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.11405819654464722, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0135, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.12952250242233276, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0119, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.09084811806678772, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0173, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.12252718955278397, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0099, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1293795108795166, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0144, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11287739127874374, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0164, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.1253224015235901, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0133, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.10046776384115219, + "learning_rate": 9.321290990574668e-06, + "loss": 0.014, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.08746109157800674, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0128, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.08306485414505005, + "learning_rate": 9.313639857014607e-06, + "loss": 0.0115, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.1340319961309433, + "learning_rate": 9.3098149378516e-06, + "loss": 0.0137, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.09587698429822922, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0143, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.14915677905082703, + "learning_rate": 9.302166396212059e-06, + "loss": 0.0143, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.09313366562128067, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0115, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.12605343759059906, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0146, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.12807077169418335, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0113, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.1186203733086586, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0119, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.10223432630300522, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0141, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.12712709605693817, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0132, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.12993593513965607, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0162, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.12273230403661728, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0133, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.1391134113073349, + "learning_rate": 9.26776941165236e-06, + "loss": 0.014, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.11695246398448944, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0137, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.1214752346277237, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0122, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.13961699604988098, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0117, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.1402655988931656, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0159, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.10683967918157578, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0146, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.08510883897542953, + "learning_rate": 9.244857665179213e-06, + "loss": 0.013, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.10308418422937393, + "learning_rate": 9.241040568803432e-06, + "loss": 0.0141, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.09736710786819458, + "learning_rate": 9.23722390978261e-06, + "loss": 0.0139, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.13270588219165802, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0118, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.08694430440664291, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0132, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.07487893849611282, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0123, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.1315941959619522, + "learning_rate": 9.221961653045449e-06, + "loss": 0.0115, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10982084274291992, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0146, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.09258668124675751, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.11609145998954773, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0143, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.1349419355392456, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0139, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.08282078802585602, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0108, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.12237592041492462, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0149, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.07864762842655182, + "learning_rate": 9.195269608854445e-06, + "loss": 0.012, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.1032051146030426, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0112, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.1346377581357956, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0132, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.08922939747571945, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0118, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.09849593788385391, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0141, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.07525688409805298, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0126, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.1287003457546234, + "learning_rate": 9.172407902473605e-06, + "loss": 0.0117, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.11016935855150223, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0123, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.09082796424627304, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0119, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.12462439388036728, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0133, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.10698465257883072, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0126, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.07600131630897522, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0133, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.09345563501119614, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0144, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.1186055988073349, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0151, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.12372780591249466, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0133, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.116783007979393, + "learning_rate": 9.138145235699377e-06, + "loss": 0.0139, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.08260510116815567, + "learning_rate": 9.134340494716546e-06, + "loss": 0.012, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.1815570890903473, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0125, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.12819825112819672, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0133, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.09238126873970032, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0115, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.09430833160877228, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0119, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.7733426094055176, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0128, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.10260160267353058, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0126, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.11146093159914017, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0136, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.11368793249130249, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0141, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.08438809961080551, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0134, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.11850261688232422, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.13420401513576508, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0128, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.09115619212388992, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0127, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.09974579513072968, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0141, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.10242673754692078, + "learning_rate": 9.081121025468816e-06, + "loss": 0.0133, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.08850608766078949, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0113, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.08285540342330933, + "learning_rate": 9.073525417675752e-06, + "loss": 0.0122, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.08347760140895844, + "learning_rate": 9.069728288749965e-06, + "loss": 0.011, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.12695780396461487, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0134, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.10132371634244919, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0143, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.07370374351739883, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0112, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.0993623211979866, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0138, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.07118473947048187, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0134, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.10804349929094315, + "learning_rate": 9.046954982950699e-06, + "loss": 0.014, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.12209299206733704, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0127, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.0896674394607544, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0134, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.10866492241621017, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0145, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.06327113509178162, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0125, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.1044771745800972, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0138, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10678671300411224, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0119, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.10146090388298035, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0122, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.11703896522521973, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0126, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.08116868138313293, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0117, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.16623066365718842, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0135, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.07943270355463028, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0128, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.17520980536937714, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0141, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.10311958938837051, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0124, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09508595615625381, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0139, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.11690590530633926, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0139, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.11585275083780289, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0119, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.08755052834749222, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0152, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.15034662187099457, + "learning_rate": 8.978732966762205e-06, + "loss": 0.013, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.11793015897274017, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0147, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.14666646718978882, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0159, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1565992683172226, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0152, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.11420148611068726, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0134, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.0945156067609787, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0105, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.0971534475684166, + "learning_rate": 8.95602513567626e-06, + "loss": 0.012, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.12953956425189972, + "learning_rate": 8.952242101985645e-06, + "loss": 0.0144, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.11144538223743439, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0138, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.11536341160535812, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0118, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.08281587064266205, + "learning_rate": 8.940895757786624e-06, + "loss": 0.0091, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.10917636007070541, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0116, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.11539111286401749, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0142, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.09897308051586151, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0141, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.12034150213003159, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0142, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.10506895184516907, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0145, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10428126901388168, + "learning_rate": 8.918215501162718e-06, + "loss": 0.013, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.0966438427567482, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0126, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.09688732028007507, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0136, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10319554805755615, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0133, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11589661985635757, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0136, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.08819710463285446, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0145, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.09449224174022675, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0141, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.11158958822488785, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0124, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.10327963531017303, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0138, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.070926733314991, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0143, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.08468566834926605, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0131, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11933926492929459, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0128, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10370025038719177, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0138, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.13985954225063324, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0144, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.08677119761705399, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0141, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.1023884266614914, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0113, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.14462946355342865, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0132, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.12994524836540222, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0144, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.08624777942895889, + "learning_rate": 8.850274743290876e-06, + "loss": 0.013, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.10459248721599579, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0117, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.12475694715976715, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0131, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09973360598087311, + "learning_rate": 8.838965950000907e-06, + "loss": 0.0142, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.10204580426216125, + "learning_rate": 8.835197287635673e-06, + "loss": 0.0136, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.08137936890125275, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0111, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.09507330507040024, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0128, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.11097851395606995, + "learning_rate": 8.823894110741103e-06, + "loss": 0.0145, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.12435153126716614, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0122, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.1186857596039772, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0123, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10177282989025116, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0119, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.08287174254655838, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0129, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.12975555658340454, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0138, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.07198113948106766, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0116, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.12089285254478455, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0126, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.09319751709699631, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0123, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.11510422080755234, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0147, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.1134580671787262, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0119, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12099597603082657, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0116, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.11170639842748642, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0111, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.0979531928896904, + "learning_rate": 8.774962556249487e-06, + "loss": 0.0137, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.0814109742641449, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0143, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.09910739213228226, + "learning_rate": 8.767441700331522e-06, + "loss": 0.0129, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.10014151781797409, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0125, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.14257559180259705, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.08456768840551376, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0138, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.19197505712509155, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0156, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.10880734771490097, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0143, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.12371200323104858, + "learning_rate": 8.744890501433196e-06, + "loss": 0.0117, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.08023890107870102, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0118, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.10231798142194748, + "learning_rate": 8.737377232363901e-06, + "loss": 0.0107, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.06945646554231644, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0131, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.09935662895441055, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0122, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.15378066897392273, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0142, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.08416267484426498, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0127, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.0862744078040123, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0132, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.11481238156557083, + "learning_rate": 8.71484884879911e-06, + "loss": 0.013, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.09441596269607544, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0126, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08091022074222565, + "learning_rate": 8.707343203095797e-06, + "loss": 0.0126, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.1480294167995453, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0121, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.12529131770133972, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0132, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.12804464995861053, + "learning_rate": 8.696088319396272e-06, + "loss": 0.013, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.10956594347953796, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0129, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.11884522438049316, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0127, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.09215931594371796, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0122, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.1411169320344925, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0139, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.07872501015663147, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0139, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.083979532122612, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0102, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.12711060047149658, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0148, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.10543761402368546, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0131, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.13125181198120117, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0142, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.13816513121128082, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0122, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.09294828027486801, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0126, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.10944795608520508, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0142, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.11008825153112411, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0122, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.10732167959213257, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0138, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.10374412685632706, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0136, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.1021609753370285, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0138, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.10304085910320282, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0125, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.08420020341873169, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0119, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.11911525577306747, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0126, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.07106399536132812, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0117, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.08642835915088654, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0125, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1176094189286232, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0146, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08892391622066498, + "learning_rate": 8.60994458849601e-06, + "loss": 0.012, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.12898282706737518, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0122, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.12732256948947906, + "learning_rate": 8.602465911610909e-06, + "loss": 0.0164, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.1723143309354782, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0146, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.10019178688526154, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0122, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.15125800669193268, + "learning_rate": 8.591251540871672e-06, + "loss": 0.0138, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.08618403971195221, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0106, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.12111509591341019, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0125, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.11654779314994812, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0133, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.12341999262571335, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0136, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.09898588061332703, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0121, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.09825480729341507, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0119, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.09357906877994537, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0125, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13970960676670074, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0131, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.14642572402954102, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0117, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.20111511647701263, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0122, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.12268098443746567, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0127, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.11043815314769745, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0109, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.12570181488990784, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0147, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.1037122830748558, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0134, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.07811714708805084, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0121, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.07688374072313309, + "learning_rate": 8.53151572917108e-06, + "loss": 0.013, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.12285313010215759, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0135, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11350435763597488, + "learning_rate": 8.524057571200438e-06, + "loss": 0.013, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.1195593997836113, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.12681595981121063, + "learning_rate": 8.516601380450119e-06, + "loss": 0.0116, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.10786957293748856, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0117, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.12382174283266068, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0123, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08464337885379791, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0139, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.11292237043380737, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0142, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.13033312559127808, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0125, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.13949616253376007, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0153, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.09846726804971695, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0115, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.12679991126060486, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0149, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.10708379000425339, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0145, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.0848502367734909, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0135, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.13396726548671722, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0142, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13248398900032043, + "learning_rate": 8.471905674372712e-06, + "loss": 0.014, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.10274169594049454, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0139, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.08507993817329407, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0112, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.10394402593374252, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0146, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.10293199867010117, + "learning_rate": 8.457022946052628e-06, + "loss": 0.015, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.16253890097141266, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0128, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.07673029601573944, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0125, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.1329779177904129, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0131, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.0754442885518074, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0125, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.09978003799915314, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0124, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.12813764810562134, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0144, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.08480092883110046, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0124, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.10068657249212265, + "learning_rate": 8.42728135848896e-06, + "loss": 0.012, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.08405598253011703, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0122, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.13025622069835663, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0117, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.11026641726493835, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0127, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.10400493443012238, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0138, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.10374867916107178, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0138, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.128962442278862, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0124, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.07732204347848892, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0124, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.14872954785823822, + "learning_rate": 8.39757171680725e-06, + "loss": 0.013, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.13792090117931366, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0175, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.07409467548131943, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0123, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.11308597028255463, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0149, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.1202300488948822, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0139, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.08481356501579285, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0124, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.07334130257368088, + "learning_rate": 8.375310536902511e-06, + "loss": 0.014, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.09884555637836456, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0137, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.16476134955883026, + "learning_rate": 8.36789416540753e-06, + "loss": 0.0142, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.12900881469249725, + "learning_rate": 8.364186735016254e-06, + "loss": 0.0138, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.10485891997814178, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0146, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.13036216795444489, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0111, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.13155606389045715, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0146, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.11586495488882065, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0135, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.1043778583407402, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0137, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.12916314601898193, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.1257125288248062, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0117, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.20916301012039185, + "learning_rate": 8.334545457592674e-06, + "loss": 0.014, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.13241995871067047, + "learning_rate": 8.330842572848802e-06, + "loss": 0.011, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.1633574515581131, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0141, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.08672885596752167, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0141, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.09803608059883118, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0143, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.13284195959568024, + "learning_rate": 8.316036101469748e-06, + "loss": 0.011, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10581238567829132, + "learning_rate": 8.312335751929666e-06, + "loss": 0.0144, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.11561308056116104, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0111, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.09897869825363159, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0116, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.129640132188797, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0142, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.08553853631019592, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.09117261320352554, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0134, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.1338709443807602, + "learning_rate": 8.290144330085352e-06, + "loss": 0.013, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.11192607134580612, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0122, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.0983581617474556, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0149, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.10853386670351028, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0145, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.09507740288972855, + "learning_rate": 8.275360236495652e-06, + "loss": 0.011, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.08634619414806366, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0146, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.12652689218521118, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0145, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.13687895238399506, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0115, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.11267609149217606, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0139, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.0999043732881546, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0125, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.09523805975914001, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0137, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09847134351730347, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0153, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.0827624723315239, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0093, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.09830755740404129, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0119, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.14614145457744598, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0149, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.1254947930574417, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0129, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.11176569014787674, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0147, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.06597425043582916, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0132, + "step": 80000 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 20000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 64, + "trial_name": null, + "trial_params": null +} diff --git a/config.json b/config.json new file mode 100644 index 0000000000000000000000000000000000000000..47d7b77e4255263f3803a4b6f94ca6831e459b70 --- /dev/null +++ b/config.json @@ -0,0 +1,64 @@ +{ + "action_dim": 32, + "action_head_cfg": { + "action_dim": 32, + "action_horizon": 16, + "add_pos_embed": true, + "backbone_embedding_dim": 2048, + "diffusion_model_cfg": { + "attention_head_dim": 48, + "cross_attention_dim": 2048, + "dropout": 0.2, + "final_dropout": true, + "interleave_self_attention": true, + "norm_type": "ada_norm", + "num_attention_heads": 32, + "num_layers": 16, + "output_dim": 1024, + "positional_embeddings": null + }, + "hidden_size": 1024, + "input_embedding_dim": 1536, + "max_action_dim": 32, + "max_state_dim": 64, + "model_dtype": "float32", + "noise_beta_alpha": 1.5, + "noise_beta_beta": 1.0, + "noise_s": 0.999, + "num_inference_timesteps": 4, + "num_target_vision_tokens": 32, + "num_timestep_buckets": 1000, + "tune_diffusion_model": true, + "tune_projector": true, + "use_vlln": true, + "vl_self_attention_cfg": { + "attention_head_dim": 64, + "dropout": 0.2, + "final_dropout": true, + "num_attention_heads": 32, + "num_layers": 4, + "positional_embeddings": null + } + }, + "action_horizon": 16, + "architectures": [ + "GR00T_N1_5" + ], + "attn_implementation": null, + "backbone_cfg": { + "eagle_path": "NVEagle/eagle_er-qwen3_1_7B-Siglip2_400M_stage1_5_128gpu_er_v7_1mlp_nops", + "load_bf16": false, + "project_to_dim": null, + "reproject_vision": false, + "select_layer": 12, + "tune_llm": false, + "tune_visual": true, + "use_flash_attention": true + }, + "compute_dtype": "bfloat16", + "hidden_size": 2048, + "model_dtype": "float32", + "model_type": "gr00t_n1_5", + "torch_dtype": "bfloat16", + "transformers_version": "4.51.3" +} diff --git a/experiment_cfg/metadata.json b/experiment_cfg/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..300eb070963337016e16465f50ba8351ea2eb25c --- /dev/null +++ b/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.3139495849609375, + 0.4876587688922882, + 0.7196521759033203 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6996827125549316 + ], + "mean": [ + 2.749288365724937, + -2.0455216239909983, + 0.700532551020533 + ], + "std": [ + 1.6786295897046262, + 1.312897969434244, + 0.00133751758906258 + ], + "q01": [ + -1.4241907881148037, + -5.1716547935950885, + 0.6999670898768614 + ], + "q99": [ + 6.198111545831555, + -0.5873531334104489, + 0.7038833014242587 + ] + }, + "base_rotation": { + "max": [ + 0.0, + 0.0, + 1.0, + 1.0 + ], + "min": [ + 0.0, + 0.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0, + 0.0, + 0.23084999587450247, + 0.6238431088581847 + ], + "std": [ + 0.0, + 0.0, + 0.6556404371644047, + 0.3572252105732042 + ], + "q01": [ + 0.0, + 0.0, + -0.9999999999999999, + 1.7482354593273349e-06 + ], + "q99": [ + 0.0, + 0.0, + 0.9999999999999999, + 0.9999999999999999 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8003877401351929, + 0.9829942584037781 + ], + "min": [ + -0.37471598386764526, + -0.8472502827644348, + -0.25070279836654663 + ], + "mean": [ + 0.28667332601863704, + -0.038315473170876704, + 0.45974658906777444 + ], + "std": [ + 0.17067058828305154, + 0.23569952037784195, + 0.21950702354181487 + ], + "q01": [ + 0.015172948963784228, + -0.44450502063220615, + 0.23314444103329338 + ], + "q99": [ + 0.5609069689572412, + 0.38454179128009547, + 0.7028102736473852 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9984455108642578, + 0.9434727430343628, + 0.9062229990959167 + ], + "min": [ + -0.9999930262565613, + -0.9988337755203247, + -0.9618149995803833, + 2.1872274658107926e-07 + ], + "mean": [ + -0.25672297401336, + 0.024425335806687216, + -0.08740977164483847, + 0.16608739644018397 + ], + "std": [ + 0.7932585208392383, + 0.3102260195580416, + 0.3772472576315148, + 0.17451959300158773 + ], + "q01": [ + -0.99379006987882, + -0.5478102050038828, + -0.6101777952971636, + 0.002274085014134748 + ], + "q99": [ + 0.8804856038896288, + 0.5910611092873037, + 0.5216058042993562, + 0.5231965974012255 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011436971835792065, + -0.05664053186774254 + ], + "mean": [ + 0.031651551516052555, + -0.03161481387920421 + ], + "std": [ + 0.013119769222217526, + 0.01306078225192402 + ], + "q01": [ + 0.0060999246446777336, + -0.04062601653964198 + ], + "q99": [ + 0.04054891621517283, + -0.0059163567113456345 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.008144896753718373, + -0.00018893135146078263, + -0.0008739062727221845, + 0.0 + ], + "std": [ + 0.11030045424364411, + 0.10148082570313594, + 0.08975698373467506, + 0.0 + ], + "q01": [ + -0.07287373067581518, + -0.09446994199569948, + -0.07702818259249572, + 0.0 + ], + "q99": [ + 0.04485485414407898, + 0.09626941605965177, + 0.07610665906122742, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9221974313425939 + ], + "std": [ + 0.38670194342612674 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.384693946883099 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.01976129923017491, + -0.020870631344490034, + -0.05993476517349181 + ], + "std": [ + 0.4347024990804053, + 0.4221662976569997, + 0.3845506843044066 + ], + "q01": [ + -0.8835792939767807, + -0.9280919251713778, + -0.783361071470956 + ], + "q99": [ + 0.7622084438449307, + 0.8625125252609077, + 0.7470974753250706 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.008089673068544335, + -0.026498254627927348, + 0.0029083551743059005 + ], + "std": [ + 0.11216734503733773, + 0.13206002613798629, + 0.12615870412692864 + ], + "q01": [ + -0.2814627972846672, + -0.4342386516115439, + -0.34489755628340574 + ], + "q99": [ + 0.31879353266939386, + 0.28411797683220563, + 0.3597718671669894 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.3824228874769045 + ], + "std": [ + 0.9240015096469786 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.5597272305813592 + ] + } + } + }, + "modalities": { + "video": { + "robot0_eye_in_hand": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_left": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + }, + "robot0_agentview_right": { + "resolution": [ + 256, + 256 + ], + "channels": 3, + "fps": 20.0 + } + }, + "state": { + "base_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "base_rotation": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "end_effector_position_relative": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation_relative": { + "absolute": true, + "rotation_type": "quaternion", + "shape": [ + 4 + ], + "continuous": true + }, + "gripper_qpos": { + "absolute": true, + "rotation_type": null, + "shape": [ + 2 + ], + "continuous": true + } + }, + "action": { + "base_motion": { + "absolute": true, + "rotation_type": null, + "shape": [ + 4 + ], + "continuous": true + }, + "control_mode": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + }, + "end_effector_position": { + "absolute": true, + "rotation_type": null, + "shape": [ + 3 + ], + "continuous": true + }, + "end_effector_rotation": { + "absolute": true, + "rotation_type": "axis_angle", + "shape": [ + 3 + ], + "continuous": true + }, + "gripper_close": { + "absolute": true, + "rotation_type": null, + "shape": [ + 1 + ], + "continuous": true + } + } + }, + "embodiment_tag": "new_embodiment" + } +} \ No newline at end of file diff --git a/model-00001-of-00002.safetensors b/model-00001-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d7d8c2b55704776186a9207b3175b6c657758e2 --- /dev/null +++ b/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9fc9945eca28b94cc6398098b3d2092b0e9731d322853d46197bf3dc7a37a961 +size 4999367032 diff --git a/model-00002-of-00002.safetensors b/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..42080be5aa6a552d612cad02b976aaea03277ce1 --- /dev/null +++ b/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:be562fdffb28a9a991de2c352b351bae7ac8e29e18c5db5b96a5fa82c26b8e39 +size 2586705312 diff --git a/model.safetensors.index.json b/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..5ef5f247b61eb93e89601c7da8334fc087e19750 --- /dev/null +++ b/model.safetensors.index.json @@ -0,0 +1,906 @@ +{ + "metadata": { + "total_size": 7585939328 + }, + "weight_map": { + "action_head.action_decoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.action_decoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W1.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W2.b": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.W": "model-00002-of-00002.safetensors", + "action_head.action_encoder.W3.b": "model-00002-of-00002.safetensors", + "action_head.future_tokens.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_1.weight": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.bias": "model-00002-of-00002.safetensors", + "action_head.model.proj_out_2.weight": "model-00002-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_1.weight": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.bias": "model-00001-of-00002.safetensors", + "action_head.model.timestep_encoder.timestep_embedder.linear_2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.0.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.1.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.10.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.11.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.12.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.13.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.14.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.bias": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.15.norm1.linear.weight": "model-00002-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.2.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.3.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.4.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.5.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.6.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.7.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.8.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_k.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_out.0.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_q.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.attn1.to_v.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.0.proj.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.ff.net.2.weight": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.bias": "model-00001-of-00002.safetensors", + "action_head.model.transformer_blocks.9.norm1.linear.weight": "model-00001-of-00002.safetensors", + "action_head.position_embedding.weight": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer1.b": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.W": "model-00002-of-00002.safetensors", + "action_head.state_encoder.layer2.b": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.0.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.1.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.2.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_k.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_out.0.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_q.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.attn1.to_v.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.0.proj.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.ff.net.2.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm1.weight": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.bias": "model-00002-of-00002.safetensors", + "action_head.vl_self_attention.transformer_blocks.3.norm3.weight": "model-00002-of-00002.safetensors", + "action_head.vlln.bias": "model-00002-of-00002.safetensors", + "action_head.vlln.weight": "model-00002-of-00002.safetensors", + "backbone.eagle_model.language_model.lm_head.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.embed_tokens.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.language_model.model.norm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.mlp1.0.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.patch_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.embeddings.position_embedding.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.layer_norm2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.in_proj_weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.attention.out_proj.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.layernorm.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc1.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.mlp.fc2.weight": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.head.probe": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.bias": "model-00001-of-00002.safetensors", + "backbone.eagle_model.vision_model.vision_model.post_layernorm.weight": "model-00001-of-00002.safetensors" + } +} diff --git a/train.log b/train.log new file mode 100644 index 0000000000000000000000000000000000000000..809a693b1c264c1ed68986355149b6d0eee04d80 --- /dev/null +++ b/train.log @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea7fa63b42f3fb5967cc4f2dbd7cb936b86ba3b4004cd9634df1013c448f7338 +size 11479468 diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..c01c0677b2abf834b55bbfc739aaa981c46a7077 --- /dev/null +++ b/trainer_state.json @@ -0,0 +1,84043 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 34.91416933372127, + "eval_steps": 500, + "global_step": 120000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.002909514111143439, + "grad_norm": 5.5908684730529785, + "learning_rate": 4.5e-08, + "loss": 0.5668, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.084368705749512, + "learning_rate": 9.5e-08, + "loss": 0.5781, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 5.755157947540283, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5759, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.174816608428955, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5202, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 6.199656963348389, + "learning_rate": 2.45e-07, + "loss": 0.5455, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.251872539520264, + "learning_rate": 2.95e-07, + "loss": 0.5507, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 5.606583595275879, + "learning_rate": 3.45e-07, + "loss": 0.5264, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.967868328094482, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.5166, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.92909574508667, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4578, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.306433200836182, + "learning_rate": 4.95e-07, + "loss": 0.4091, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.75846529006958, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4072, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.2030856609344482, + "learning_rate": 5.95e-07, + "loss": 0.2992, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.5756216049194336, + "learning_rate": 6.45e-07, + "loss": 0.2703, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.3420923948287964, + "learning_rate": 6.95e-07, + "loss": 0.2492, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 0.9191234111785889, + "learning_rate": 7.45e-07, + "loss": 0.2302, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.662022054195404, + "learning_rate": 7.95e-07, + "loss": 0.1993, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.704409658908844, + "learning_rate": 8.45e-07, + "loss": 0.1814, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.6104884147644043, + "learning_rate": 8.95e-07, + "loss": 0.1694, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.4955693781375885, + "learning_rate": 9.450000000000001e-07, + "loss": 0.1674, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.707667887210846, + "learning_rate": 9.95e-07, + "loss": 0.1507, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.6977017521858215, + "learning_rate": 1.045e-06, + "loss": 0.1547, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.49085867404937744, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1536, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.6363996267318726, + "learning_rate": 1.145e-06, + "loss": 0.1565, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.4633055627346039, + "learning_rate": 1.195e-06, + "loss": 0.1641, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5155240297317505, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1513, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5815547108650208, + "learning_rate": 1.295e-06, + "loss": 0.1431, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.5865835547447205, + "learning_rate": 1.345e-06, + "loss": 0.1419, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.5872474312782288, + "learning_rate": 1.395e-06, + "loss": 0.1326, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.4897519052028656, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1326, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.46316635608673096, + "learning_rate": 1.495e-06, + "loss": 0.1442, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.5359047651290894, + "learning_rate": 1.545e-06, + "loss": 0.1394, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.5805708169937134, + "learning_rate": 1.595e-06, + "loss": 0.1231, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.5488415360450745, + "learning_rate": 1.645e-06, + "loss": 0.127, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.41755908727645874, + "learning_rate": 1.695e-06, + "loss": 0.1375, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.6258692145347595, + "learning_rate": 1.745e-06, + "loss": 0.1202, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.5017617344856262, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1317, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.710078239440918, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1207, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.46977895498275757, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1215, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.4802077114582062, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.116, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.6448793411254883, + "learning_rate": 1.995e-06, + "loss": 0.1172, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4394509494304657, + "learning_rate": 2.045e-06, + "loss": 0.1145, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.6547806859016418, + "learning_rate": 2.095e-06, + "loss": 0.1174, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.6954914331436157, + "learning_rate": 2.145e-06, + "loss": 0.1245, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.670100212097168, + "learning_rate": 2.195e-06, + "loss": 0.1133, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.5230806469917297, + "learning_rate": 2.245e-06, + "loss": 0.11, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.46809589862823486, + "learning_rate": 2.295e-06, + "loss": 0.1069, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.6178835034370422, + "learning_rate": 2.345e-06, + "loss": 0.1099, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.41567298769950867, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1072, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4968737065792084, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1093, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.46893924474716187, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1046, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.5041665434837341, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.1014, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5833673477172852, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1095, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.5565881133079529, + "learning_rate": 2.645e-06, + "loss": 0.1057, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6213339567184448, + "learning_rate": 2.695e-06, + "loss": 0.1087, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.5638794898986816, + "learning_rate": 2.745e-06, + "loss": 0.099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.5594047904014587, + "learning_rate": 2.795e-06, + "loss": 0.1065, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5521039366722107, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1072, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.5013270378112793, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.1045, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.44355231523513794, + "learning_rate": 2.945e-06, + "loss": 0.0985, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.5467342734336853, + "learning_rate": 2.995e-06, + "loss": 0.101, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5065611600875854, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0939, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.5587425827980042, + "learning_rate": 3.095e-06, + "loss": 0.0969, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5561433434486389, + "learning_rate": 3.145e-06, + "loss": 0.1009, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.5576680302619934, + "learning_rate": 3.195e-06, + "loss": 0.1006, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.7426344752311707, + "learning_rate": 3.245e-06, + "loss": 0.0959, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.64913409948349, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.0962, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.5271337628364563, + "learning_rate": 3.345e-06, + "loss": 0.0927, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.6992810368537903, + "learning_rate": 3.395e-06, + "loss": 0.0924, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.559481680393219, + "learning_rate": 3.445e-06, + "loss": 0.0886, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7303935885429382, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0918, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.45654281973838806, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0934, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.41098082065582275, + "learning_rate": 3.595e-06, + "loss": 0.0911, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.43506675958633423, + "learning_rate": 3.645e-06, + "loss": 0.0887, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5131197571754456, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0983, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.4571124017238617, + "learning_rate": 3.745e-06, + "loss": 0.0928, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.6749465465545654, + "learning_rate": 3.795e-06, + "loss": 0.0926, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.46945369243621826, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.1016, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.6481541395187378, + "learning_rate": 3.895e-06, + "loss": 0.0917, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.6769195795059204, + "learning_rate": 3.945e-06, + "loss": 0.0924, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5225585103034973, + "learning_rate": 3.995e-06, + "loss": 0.0909, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.642867922782898, + "learning_rate": 4.045e-06, + "loss": 0.0829, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.4617709219455719, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0924, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.5350925326347351, + "learning_rate": 4.145e-06, + "loss": 0.0821, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.4871961176395416, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0896, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.5545480847358704, + "learning_rate": 4.245e-06, + "loss": 0.0891, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.6057930588722229, + "learning_rate": 4.295e-06, + "loss": 0.0876, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.515282154083252, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0873, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.5402254462242126, + "learning_rate": 4.395e-06, + "loss": 0.0908, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4669134318828583, + "learning_rate": 4.445e-06, + "loss": 0.0837, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.46819189190864563, + "learning_rate": 4.495e-06, + "loss": 0.0888, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35766035318374634, + "learning_rate": 4.545e-06, + "loss": 0.0829, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.6769008040428162, + "learning_rate": 4.595e-06, + "loss": 0.0817, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5884619355201721, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0892, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5613101720809937, + "learning_rate": 4.695e-06, + "loss": 0.0852, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.47087088227272034, + "learning_rate": 4.745e-06, + "loss": 0.0818, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.4887458384037018, + "learning_rate": 4.795e-06, + "loss": 0.0858, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.44256526231765747, + "learning_rate": 4.845e-06, + "loss": 0.0833, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5375688076019287, + "learning_rate": 4.895e-06, + "loss": 0.075, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.6297019124031067, + "learning_rate": 4.945e-06, + "loss": 0.0932, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.6576125621795654, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0829, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.724773108959198, + "learning_rate": 5.045e-06, + "loss": 0.0813, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.54575115442276, + "learning_rate": 5.095e-06, + "loss": 0.0873, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.5186223387718201, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0736, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.46651381254196167, + "learning_rate": 5.195e-06, + "loss": 0.0791, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.5323564410209656, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0753, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.42579638957977295, + "learning_rate": 5.295e-06, + "loss": 0.0755, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.5063076019287109, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.0767, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.4558819532394409, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0807, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.6920697093009949, + "learning_rate": 5.445e-06, + "loss": 0.079, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.4706108868122101, + "learning_rate": 5.495e-06, + "loss": 0.079, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.46705493330955505, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0745, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4643767774105072, + "learning_rate": 5.595e-06, + "loss": 0.0778, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.4800759553909302, + "learning_rate": 5.645e-06, + "loss": 0.0792, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.37705323100090027, + "learning_rate": 5.695e-06, + "loss": 0.0815, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.6651142239570618, + "learning_rate": 5.745e-06, + "loss": 0.0732, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.5275036692619324, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0758, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.45047008991241455, + "learning_rate": 5.845e-06, + "loss": 0.0808, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.6931971311569214, + "learning_rate": 5.895e-06, + "loss": 0.0808, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.7053916454315186, + "learning_rate": 5.945e-06, + "loss": 0.0747, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.6798260807991028, + "learning_rate": 5.995e-06, + "loss": 0.0829, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.5724003314971924, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0749, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.5836488008499146, + "learning_rate": 6.095e-06, + "loss": 0.0759, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.40817850828170776, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0778, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.41070470213890076, + "learning_rate": 6.195e-06, + "loss": 0.0765, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.43642961978912354, + "learning_rate": 6.245e-06, + "loss": 0.0711, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.49526578187942505, + "learning_rate": 6.295000000000001e-06, + "loss": 0.0713, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.7773705720901489, + "learning_rate": 6.345e-06, + "loss": 0.0744, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.5485936999320984, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0732, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.5593521595001221, + "learning_rate": 6.445e-06, + "loss": 0.0747, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.3667270839214325, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0723, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.5555917620658875, + "learning_rate": 6.545e-06, + "loss": 0.0726, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.38107141852378845, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0647, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.46452224254608154, + "learning_rate": 6.645e-06, + "loss": 0.0787, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6018183827400208, + "learning_rate": 6.695e-06, + "loss": 0.0718, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.44997578859329224, + "learning_rate": 6.745e-06, + "loss": 0.0758, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45717573165893555, + "learning_rate": 6.795e-06, + "loss": 0.0738, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.5185182690620422, + "learning_rate": 6.845e-06, + "loss": 0.0768, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.5135819315910339, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.43529245257377625, + "learning_rate": 6.945e-06, + "loss": 0.0691, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.49389752745628357, + "learning_rate": 6.995e-06, + "loss": 0.0678, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.4113156199455261, + "learning_rate": 7.045e-06, + "loss": 0.0651, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.4072611927986145, + "learning_rate": 7.095e-06, + "loss": 0.0691, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.4524228870868683, + "learning_rate": 7.145e-06, + "loss": 0.0688, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.6225605607032776, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.0676, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4841763377189636, + "learning_rate": 7.245e-06, + "loss": 0.0678, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.46297937631607056, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0759, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.5585324168205261, + "learning_rate": 7.345e-06, + "loss": 0.0671, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.57328200340271, + "learning_rate": 7.395e-06, + "loss": 0.0717, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.5072467923164368, + "learning_rate": 7.445000000000001e-06, + "loss": 0.071, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.4630240797996521, + "learning_rate": 7.495e-06, + "loss": 0.0674, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.35908272862434387, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0704, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.5117342472076416, + "learning_rate": 7.595e-06, + "loss": 0.0675, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.5214630961418152, + "learning_rate": 7.645e-06, + "loss": 0.0702, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.39036089181900024, + "learning_rate": 7.695e-06, + "loss": 0.0709, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.558722734451294, + "learning_rate": 7.745e-06, + "loss": 0.0705, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.3612176179885864, + "learning_rate": 7.795e-06, + "loss": 0.0715, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.523444652557373, + "learning_rate": 7.845e-06, + "loss": 0.0685, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.5231291651725769, + "learning_rate": 7.895e-06, + "loss": 0.0717, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.50368332862854, + "learning_rate": 7.945e-06, + "loss": 0.0694, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.4126806855201721, + "learning_rate": 7.995e-06, + "loss": 0.0693, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.719352662563324, + "learning_rate": 8.045e-06, + "loss": 0.0651, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.4405677020549774, + "learning_rate": 8.095e-06, + "loss": 0.0645, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.34746286273002625, + "learning_rate": 8.145e-06, + "loss": 0.0687, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.43200597167015076, + "learning_rate": 8.195e-06, + "loss": 0.0655, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.3464098274707794, + "learning_rate": 8.245e-06, + "loss": 0.0639, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.27263233065605164, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0629, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.343181848526001, + "learning_rate": 8.345e-06, + "loss": 0.0695, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3949901759624481, + "learning_rate": 8.395e-06, + "loss": 0.0643, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.4489574730396271, + "learning_rate": 8.445e-06, + "loss": 0.0698, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.4746992588043213, + "learning_rate": 8.495e-06, + "loss": 0.0702, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.3651655316352844, + "learning_rate": 8.545e-06, + "loss": 0.0673, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.471071720123291, + "learning_rate": 8.595e-06, + "loss": 0.0648, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.35623008012771606, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0698, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.41837432980537415, + "learning_rate": 8.695e-06, + "loss": 0.0639, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.45608940720558167, + "learning_rate": 8.745e-06, + "loss": 0.0707, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.40346869826316833, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0625, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.479824960231781, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0703, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.38570791482925415, + "learning_rate": 8.895e-06, + "loss": 0.0643, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.36530473828315735, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0652, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37868866324424744, + "learning_rate": 8.995000000000001e-06, + "loss": 0.068, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.32546791434288025, + "learning_rate": 9.045e-06, + "loss": 0.0636, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4664970338344574, + "learning_rate": 9.095e-06, + "loss": 0.0671, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4767569601535797, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0625, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.41823142766952515, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0685, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.3709985613822937, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0681, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.3933296799659729, + "learning_rate": 9.295e-06, + "loss": 0.0678, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.3079732060432434, + "learning_rate": 9.345e-06, + "loss": 0.0598, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.33485978841781616, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0644, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.40312913060188293, + "learning_rate": 9.445e-06, + "loss": 0.0684, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.4334392547607422, + "learning_rate": 9.495e-06, + "loss": 0.0622, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.5787541270256042, + "learning_rate": 9.545e-06, + "loss": 0.0656, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.2827294170856476, + "learning_rate": 9.595e-06, + "loss": 0.0628, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4431678056716919, + "learning_rate": 9.645e-06, + "loss": 0.0633, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.3916102647781372, + "learning_rate": 9.695e-06, + "loss": 0.0632, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.41470348834991455, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0612, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.3224067986011505, + "learning_rate": 9.795e-06, + "loss": 0.061, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.4110186696052551, + "learning_rate": 9.845e-06, + "loss": 0.066, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.4786360561847687, + "learning_rate": 9.895e-06, + "loss": 0.0648, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3667179048061371, + "learning_rate": 9.945e-06, + "loss": 0.0663, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.31722185015678406, + "learning_rate": 9.995e-06, + "loss": 0.0588, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.29271745681762695, + "learning_rate": 1.0045e-05, + "loss": 0.0636, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.3536592423915863, + "learning_rate": 1.0095e-05, + "loss": 0.0625, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.3228510916233063, + "learning_rate": 1.0145e-05, + "loss": 0.067, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.38474392890930176, + "learning_rate": 1.0195e-05, + "loss": 0.0609, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.40410760045051575, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0638, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.3632320463657379, + "learning_rate": 1.0295e-05, + "loss": 0.0587, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.37775692343711853, + "learning_rate": 1.0345e-05, + "loss": 0.0673, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.3343006372451782, + "learning_rate": 1.0395e-05, + "loss": 0.0599, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3232511281967163, + "learning_rate": 1.0445e-05, + "loss": 0.0661, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.38730430603027344, + "learning_rate": 1.0495e-05, + "loss": 0.0649, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.38972100615501404, + "learning_rate": 1.0545e-05, + "loss": 0.0582, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.27242428064346313, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0583, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.30514073371887207, + "learning_rate": 1.0645e-05, + "loss": 0.065, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.45364341139793396, + "learning_rate": 1.0695e-05, + "loss": 0.0576, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.3863564729690552, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0579, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.33717092871665955, + "learning_rate": 1.0795e-05, + "loss": 0.0565, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38073790073394775, + "learning_rate": 1.0845e-05, + "loss": 0.0585, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4277617037296295, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0622, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.3545698821544647, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0609, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.30646640062332153, + "learning_rate": 1.0995e-05, + "loss": 0.0659, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.3607352375984192, + "learning_rate": 1.1045e-05, + "loss": 0.0621, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3624425530433655, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.0626, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3856692612171173, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0604, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.4390200972557068, + "learning_rate": 1.1195e-05, + "loss": 0.0601, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.4108872711658478, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.0618, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.32732832431793213, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0636, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.3602239787578583, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.29978248476982117, + "learning_rate": 1.1395e-05, + "loss": 0.058, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.48490920662879944, + "learning_rate": 1.1445e-05, + "loss": 0.0614, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.3278113901615143, + "learning_rate": 1.1495e-05, + "loss": 0.0611, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.37862712144851685, + "learning_rate": 1.1545e-05, + "loss": 0.0626, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.3514714241027832, + "learning_rate": 1.1595e-05, + "loss": 0.0583, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.27256545424461365, + "learning_rate": 1.1645e-05, + "loss": 0.062, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.3305833339691162, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0572, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.2881031930446625, + "learning_rate": 1.1745e-05, + "loss": 0.0641, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.3219377398490906, + "learning_rate": 1.1795e-05, + "loss": 0.0603, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.34189504384994507, + "learning_rate": 1.1845e-05, + "loss": 0.0607, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30541297793388367, + "learning_rate": 1.1895e-05, + "loss": 0.056, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.35898879170417786, + "learning_rate": 1.1945e-05, + "loss": 0.0598, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.33221277594566345, + "learning_rate": 1.1995e-05, + "loss": 0.0603, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.3435590863227844, + "learning_rate": 1.2045e-05, + "loss": 0.0553, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.24820426106452942, + "learning_rate": 1.2095e-05, + "loss": 0.0644, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.3749466836452484, + "learning_rate": 1.2145e-05, + "loss": 0.0615, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.3984697759151459, + "learning_rate": 1.2195e-05, + "loss": 0.0595, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3250048756599426, + "learning_rate": 1.2245e-05, + "loss": 0.0608, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.2794302701950073, + "learning_rate": 1.2295e-05, + "loss": 0.0578, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3811834752559662, + "learning_rate": 1.2345e-05, + "loss": 0.0612, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3160156309604645, + "learning_rate": 1.2395e-05, + "loss": 0.0541, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.3825189173221588, + "learning_rate": 1.2445e-05, + "loss": 0.0621, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.3966415226459503, + "learning_rate": 1.2495e-05, + "loss": 0.0598, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3615199625492096, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0616, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.440762996673584, + "learning_rate": 1.2595e-05, + "loss": 0.0599, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.345599889755249, + "learning_rate": 1.2645e-05, + "loss": 0.0597, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.2963278889656067, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0556, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.3720734417438507, + "learning_rate": 1.2745e-05, + "loss": 0.0618, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3937445878982544, + "learning_rate": 1.2795e-05, + "loss": 0.0624, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.4492146372795105, + "learning_rate": 1.2845e-05, + "loss": 0.0573, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.5172158479690552, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0594, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.3796320855617523, + "learning_rate": 1.2945e-05, + "loss": 0.0528, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.42423033714294434, + "learning_rate": 1.2995e-05, + "loss": 0.0615, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.4578545093536377, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0562, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.38345491886138916, + "learning_rate": 1.3095e-05, + "loss": 0.0574, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.2539816200733185, + "learning_rate": 1.3145e-05, + "loss": 0.0562, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.44323891401290894, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.0588, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.4074377417564392, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0591, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.5252887606620789, + "learning_rate": 1.3295e-05, + "loss": 0.0583, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.39536044001579285, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0588, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.6300645470619202, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0584, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.312466561794281, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0588, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3810383379459381, + "learning_rate": 1.3495e-05, + "loss": 0.0597, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.30645400285720825, + "learning_rate": 1.3545e-05, + "loss": 0.0589, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.2857605218887329, + "learning_rate": 1.3595e-05, + "loss": 0.0589, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.3206622898578644, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0597, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3226073086261749, + "learning_rate": 1.3695e-05, + "loss": 0.0559, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.24912619590759277, + "learning_rate": 1.3745e-05, + "loss": 0.0547, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.29700762033462524, + "learning_rate": 1.3795e-05, + "loss": 0.0563, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.32627588510513306, + "learning_rate": 1.3845e-05, + "loss": 0.0602, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.48569780588150024, + "learning_rate": 1.3895e-05, + "loss": 0.0553, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.323607474565506, + "learning_rate": 1.3945e-05, + "loss": 0.0578, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.42626142501831055, + "learning_rate": 1.3995e-05, + "loss": 0.0539, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.3281888961791992, + "learning_rate": 1.4045e-05, + "loss": 0.0638, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.38216397166252136, + "learning_rate": 1.4095e-05, + "loss": 0.0572, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.2816051244735718, + "learning_rate": 1.4145e-05, + "loss": 0.0543, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.2604160010814667, + "learning_rate": 1.4195e-05, + "loss": 0.0555, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.34318771958351135, + "learning_rate": 1.4245e-05, + "loss": 0.0565, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.31248190999031067, + "learning_rate": 1.4295e-05, + "loss": 0.0583, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.3014426827430725, + "learning_rate": 1.4345e-05, + "loss": 0.0586, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3650360703468323, + "learning_rate": 1.4395e-05, + "loss": 0.0581, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.33519116044044495, + "learning_rate": 1.4445e-05, + "loss": 0.0611, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.3595008850097656, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.054, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.32799217104911804, + "learning_rate": 1.4545e-05, + "loss": 0.0574, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.29410457611083984, + "learning_rate": 1.4595e-05, + "loss": 0.0582, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.2727576792240143, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0588, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.36562401056289673, + "learning_rate": 1.4695e-05, + "loss": 0.0597, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.3351168930530548, + "learning_rate": 1.4745e-05, + "loss": 0.0558, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.23880337178707123, + "learning_rate": 1.4795e-05, + "loss": 0.0571, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.3910059928894043, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0573, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.3103587031364441, + "learning_rate": 1.4895e-05, + "loss": 0.0578, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.3265620768070221, + "learning_rate": 1.4945e-05, + "loss": 0.0542, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.3060731291770935, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0537, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.41589322686195374, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0546, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.3102424740791321, + "learning_rate": 1.5095e-05, + "loss": 0.0564, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.32286810874938965, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0519, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.3129369020462036, + "learning_rate": 1.5195e-05, + "loss": 0.0566, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.43735018372535706, + "learning_rate": 1.5245e-05, + "loss": 0.0552, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.440449595451355, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0589, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.3631616532802582, + "learning_rate": 1.5345e-05, + "loss": 0.0511, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.29124221205711365, + "learning_rate": 1.5395e-05, + "loss": 0.055, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.4247622489929199, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0562, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.36976534128189087, + "learning_rate": 1.5495e-05, + "loss": 0.0604, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.3580514192581177, + "learning_rate": 1.5545e-05, + "loss": 0.0547, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.2969493567943573, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.0564, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28583595156669617, + "learning_rate": 1.5645e-05, + "loss": 0.0586, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.3220328390598297, + "learning_rate": 1.5695e-05, + "loss": 0.0588, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.32270652055740356, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0559, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.3170712888240814, + "learning_rate": 1.5795e-05, + "loss": 0.0566, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.2966325879096985, + "learning_rate": 1.5845e-05, + "loss": 0.0551, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.2788766622543335, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.051, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.3521708548069, + "learning_rate": 1.5945e-05, + "loss": 0.0558, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.35105040669441223, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0551, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.4221852123737335, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0573, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.4217511713504791, + "learning_rate": 1.6095e-05, + "loss": 0.0532, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.4519175589084625, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0489, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3427059054374695, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0544, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29204127192497253, + "learning_rate": 1.6245e-05, + "loss": 0.0525, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.24138635396957397, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0559, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.3599991500377655, + "learning_rate": 1.6345e-05, + "loss": 0.0544, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.3326258361339569, + "learning_rate": 1.6395e-05, + "loss": 0.0593, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31976261734962463, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0624, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.26379555463790894, + "learning_rate": 1.6495e-05, + "loss": 0.0529, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2576923966407776, + "learning_rate": 1.6545e-05, + "loss": 0.0527, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.30890268087387085, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.0571, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.4547998309135437, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.3050350248813629, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0552, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.2984176278114319, + "learning_rate": 1.6745e-05, + "loss": 0.0528, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.34583792090415955, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0512, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.33085328340530396, + "learning_rate": 1.6845e-05, + "loss": 0.0563, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.33391326665878296, + "learning_rate": 1.6895e-05, + "loss": 0.0527, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.23657914996147156, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0509, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.3719659149646759, + "learning_rate": 1.6995e-05, + "loss": 0.0568, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.4268389940261841, + "learning_rate": 1.7045e-05, + "loss": 0.0525, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.2726958394050598, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0482, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27847009897232056, + "learning_rate": 1.7145e-05, + "loss": 0.0559, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.26195716857910156, + "learning_rate": 1.7195e-05, + "loss": 0.0588, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.2666028141975403, + "learning_rate": 1.7245e-05, + "loss": 0.0548, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.26048436760902405, + "learning_rate": 1.7295e-05, + "loss": 0.0592, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.22569318115711212, + "learning_rate": 1.7345e-05, + "loss": 0.0505, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.35350096225738525, + "learning_rate": 1.7395e-05, + "loss": 0.0566, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.3166314363479614, + "learning_rate": 1.7445e-05, + "loss": 0.0517, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.24918541312217712, + "learning_rate": 1.7495e-05, + "loss": 0.0545, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3416207730770111, + "learning_rate": 1.7545e-05, + "loss": 0.0544, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.22633257508277893, + "learning_rate": 1.7595e-05, + "loss": 0.0545, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.3972736597061157, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0591, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3213656544685364, + "learning_rate": 1.7695e-05, + "loss": 0.0537, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.30361926555633545, + "learning_rate": 1.7745e-05, + "loss": 0.0529, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.35521000623703003, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0549, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.4444983899593353, + "learning_rate": 1.7845e-05, + "loss": 0.0529, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.32495826482772827, + "learning_rate": 1.7895e-05, + "loss": 0.0557, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.2893896698951721, + "learning_rate": 1.7945e-05, + "loss": 0.056, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.2607886493206024, + "learning_rate": 1.7995e-05, + "loss": 0.055, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.3305937647819519, + "learning_rate": 1.8045e-05, + "loss": 0.0529, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.3003319799900055, + "learning_rate": 1.8095e-05, + "loss": 0.0515, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3523712456226349, + "learning_rate": 1.8145e-05, + "loss": 0.0524, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.27575406432151794, + "learning_rate": 1.8195e-05, + "loss": 0.0535, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.36445286870002747, + "learning_rate": 1.8245e-05, + "loss": 0.0552, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.41415491700172424, + "learning_rate": 1.8295e-05, + "loss": 0.0537, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.3738953471183777, + "learning_rate": 1.8345e-05, + "loss": 0.0555, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.3446105420589447, + "learning_rate": 1.8395e-05, + "loss": 0.0526, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.39016082882881165, + "learning_rate": 1.8445e-05, + "loss": 0.0536, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.4075721502304077, + "learning_rate": 1.8495e-05, + "loss": 0.0548, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.4071345925331116, + "learning_rate": 1.8545e-05, + "loss": 0.0539, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.34446755051612854, + "learning_rate": 1.8595e-05, + "loss": 0.057, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.3431609272956848, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.0515, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.31537994742393494, + "learning_rate": 1.8695e-05, + "loss": 0.0534, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.4846970736980438, + "learning_rate": 1.8745e-05, + "loss": 0.0532, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.3109683394432068, + "learning_rate": 1.8795e-05, + "loss": 0.0547, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.28903284668922424, + "learning_rate": 1.8845e-05, + "loss": 0.0505, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.26592350006103516, + "learning_rate": 1.8895e-05, + "loss": 0.0499, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.27347302436828613, + "learning_rate": 1.8945e-05, + "loss": 0.052, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.3194405138492584, + "learning_rate": 1.8995e-05, + "loss": 0.0521, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.40457430481910706, + "learning_rate": 1.9045e-05, + "loss": 0.0523, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.3055919408798218, + "learning_rate": 1.9095e-05, + "loss": 0.0504, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.4911873936653137, + "learning_rate": 1.9145e-05, + "loss": 0.0531, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.30619072914123535, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0525, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.24404728412628174, + "learning_rate": 1.9245e-05, + "loss": 0.0498, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.31723177433013916, + "learning_rate": 1.9295e-05, + "loss": 0.0519, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.28105342388153076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0528, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.32965341210365295, + "learning_rate": 1.9395e-05, + "loss": 0.0504, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.2487417608499527, + "learning_rate": 1.9445e-05, + "loss": 0.0552, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3788810670375824, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0535, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.2561226189136505, + "learning_rate": 1.9545e-05, + "loss": 0.056, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.29964715242385864, + "learning_rate": 1.9595e-05, + "loss": 0.0565, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.386600136756897, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0468, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.37690475583076477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.35367539525032043, + "learning_rate": 1.9745e-05, + "loss": 0.0501, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.30189844965934753, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0524, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.27525416016578674, + "learning_rate": 1.9845e-05, + "loss": 0.0524, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.22557462751865387, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0593, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.3572654128074646, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0536, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.37130483984947205, + "learning_rate": 1.9995e-05, + "loss": 0.0507, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.40820422768592834, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0547, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.24222403764724731, + "learning_rate": 2.0095e-05, + "loss": 0.0536, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.35473307967185974, + "learning_rate": 2.0145e-05, + "loss": 0.0503, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24238434433937073, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0511, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.32721346616744995, + "learning_rate": 2.0245e-05, + "loss": 0.0474, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.3787407875061035, + "learning_rate": 2.0295e-05, + "loss": 0.0515, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.3435424566268921, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0521, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.250515341758728, + "learning_rate": 2.0395e-05, + "loss": 0.0539, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.2730806767940521, + "learning_rate": 2.0445e-05, + "loss": 0.0544, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.32704228162765503, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0516, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.26865822076797485, + "learning_rate": 2.0545e-05, + "loss": 0.0557, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.31514087319374084, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.051, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.2616795301437378, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0483, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2979176938533783, + "learning_rate": 2.0695e-05, + "loss": 0.056, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.2876877188682556, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0496, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.2801620662212372, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0533, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.3521801233291626, + "learning_rate": 2.0845e-05, + "loss": 0.051, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.3016152083873749, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.0515, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.29837727546691895, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0483, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.255982369184494, + "learning_rate": 2.0995e-05, + "loss": 0.0482, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.20852255821228027, + "learning_rate": 2.1045e-05, + "loss": 0.05, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.2746782898902893, + "learning_rate": 2.1095e-05, + "loss": 0.0513, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.28745001554489136, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0486, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.3163246810436249, + "learning_rate": 2.1195e-05, + "loss": 0.0536, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.2279432713985443, + "learning_rate": 2.1245e-05, + "loss": 0.0497, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.2273164689540863, + "learning_rate": 2.1295e-05, + "loss": 0.0485, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.3124007284641266, + "learning_rate": 2.1345e-05, + "loss": 0.0473, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.2097717970609665, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0514, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.24428582191467285, + "learning_rate": 2.1445e-05, + "loss": 0.0543, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.47236987948417664, + "learning_rate": 2.1495e-05, + "loss": 0.0525, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.3367164433002472, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0525, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.34086376428604126, + "learning_rate": 2.1595e-05, + "loss": 0.049, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.3060646951198578, + "learning_rate": 2.1645e-05, + "loss": 0.0538, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.2997030019760132, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0553, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.2652549743652344, + "learning_rate": 2.1745e-05, + "loss": 0.051, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.3072361350059509, + "learning_rate": 2.1795e-05, + "loss": 0.0516, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2494215965270996, + "learning_rate": 2.1845e-05, + "loss": 0.0476, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.20796196162700653, + "learning_rate": 2.1895e-05, + "loss": 0.0553, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.24235095083713531, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.21033242344856262, + "learning_rate": 2.1995e-05, + "loss": 0.0542, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.26056164503097534, + "learning_rate": 2.2045e-05, + "loss": 0.0546, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.24798662960529327, + "learning_rate": 2.2095e-05, + "loss": 0.0529, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.22179707884788513, + "learning_rate": 2.2145e-05, + "loss": 0.0542, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.2660025358200073, + "learning_rate": 2.2195e-05, + "loss": 0.0496, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.3771500289440155, + "learning_rate": 2.2245e-05, + "loss": 0.0502, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.3059975206851959, + "learning_rate": 2.2295e-05, + "loss": 0.0482, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.2941029667854309, + "learning_rate": 2.2345e-05, + "loss": 0.0483, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3481833040714264, + "learning_rate": 2.2395e-05, + "loss": 0.0541, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.2729704678058624, + "learning_rate": 2.2445e-05, + "loss": 0.0514, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.44202038645744324, + "learning_rate": 2.2495e-05, + "loss": 0.0534, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.311088889837265, + "learning_rate": 2.2545e-05, + "loss": 0.0474, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30663055181503296, + "learning_rate": 2.2595e-05, + "loss": 0.0491, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.3530942499637604, + "learning_rate": 2.2645e-05, + "loss": 0.0503, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.27998727560043335, + "learning_rate": 2.2695e-05, + "loss": 0.0493, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.2679221034049988, + "learning_rate": 2.2745e-05, + "loss": 0.0512, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.348821222782135, + "learning_rate": 2.2795e-05, + "loss": 0.0489, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.260958194732666, + "learning_rate": 2.2845e-05, + "loss": 0.0474, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.2429468184709549, + "learning_rate": 2.2895e-05, + "loss": 0.053, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.2635575234889984, + "learning_rate": 2.2945e-05, + "loss": 0.053, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.3416514992713928, + "learning_rate": 2.2995e-05, + "loss": 0.0476, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.221172496676445, + "learning_rate": 2.3045e-05, + "loss": 0.0519, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.30541878938674927, + "learning_rate": 2.3095e-05, + "loss": 0.0522, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.25274574756622314, + "learning_rate": 2.3145e-05, + "loss": 0.0499, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.366253525018692, + "learning_rate": 2.3195e-05, + "loss": 0.0463, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.16825024783611298, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0493, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.3158467710018158, + "learning_rate": 2.3295e-05, + "loss": 0.0482, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.2887287139892578, + "learning_rate": 2.3345e-05, + "loss": 0.053, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.4717058539390564, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0515, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.3148113489151001, + "learning_rate": 2.3445e-05, + "loss": 0.0523, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.23815515637397766, + "learning_rate": 2.3495e-05, + "loss": 0.0523, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23897255957126617, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0484, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.2680213749408722, + "learning_rate": 2.3595e-05, + "loss": 0.0485, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.29999756813049316, + "learning_rate": 2.3645e-05, + "loss": 0.0506, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.30607151985168457, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.05, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.26077136397361755, + "learning_rate": 2.3745e-05, + "loss": 0.0558, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2638765275478363, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.049, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.24569877982139587, + "learning_rate": 2.3845e-05, + "loss": 0.053, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.29622989892959595, + "learning_rate": 2.3895e-05, + "loss": 0.0517, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.2598831355571747, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.05, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.25716227293014526, + "learning_rate": 2.3995e-05, + "loss": 0.0453, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.37305381894111633, + "learning_rate": 2.4045e-05, + "loss": 0.0539, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.3151630461215973, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0516, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.296941339969635, + "learning_rate": 2.4145e-05, + "loss": 0.0509, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.31995704770088196, + "learning_rate": 2.4195e-05, + "loss": 0.0498, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2426208257675171, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0475, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.23927678167819977, + "learning_rate": 2.4295e-05, + "loss": 0.05, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.1887904703617096, + "learning_rate": 2.4345e-05, + "loss": 0.05, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.18480589985847473, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0496, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.24515867233276367, + "learning_rate": 2.4445e-05, + "loss": 0.0464, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.25505101680755615, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0481, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.24435125291347504, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0502, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.25899314880371094, + "learning_rate": 2.4595e-05, + "loss": 0.0535, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.30518922209739685, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0481, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.26098036766052246, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.0477, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.24723340570926666, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.26713746786117554, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.0506, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.21166373789310455, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0502, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2602159082889557, + "learning_rate": 2.4895e-05, + "loss": 0.0506, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27918240427970886, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.049, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.2131885290145874, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0477, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.28778040409088135, + "learning_rate": 2.5045e-05, + "loss": 0.0487, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.3056010603904724, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.049, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.2850024700164795, + "learning_rate": 2.5145e-05, + "loss": 0.0475, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.20490600168704987, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0497, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.32371339201927185, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0491, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.32466208934783936, + "learning_rate": 2.5295e-05, + "loss": 0.0485, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.2779853343963623, + "learning_rate": 2.5345e-05, + "loss": 0.0485, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.3257012367248535, + "learning_rate": 2.5395e-05, + "loss": 0.0486, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2625649571418762, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0486, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.19846542179584503, + "learning_rate": 2.5495e-05, + "loss": 0.0494, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.2415613830089569, + "learning_rate": 2.5545e-05, + "loss": 0.0539, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.20438073575496674, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0476, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.2543872892856598, + "learning_rate": 2.5645e-05, + "loss": 0.0489, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.26969635486602783, + "learning_rate": 2.5695e-05, + "loss": 0.0557, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.27818888425827026, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0517, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.28826993703842163, + "learning_rate": 2.5795e-05, + "loss": 0.0499, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.19518250226974487, + "learning_rate": 2.5845e-05, + "loss": 0.0499, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.2639068067073822, + "learning_rate": 2.5895e-05, + "loss": 0.0459, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.26014742255210876, + "learning_rate": 2.5945e-05, + "loss": 0.0522, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.26659756898880005, + "learning_rate": 2.5995e-05, + "loss": 0.0471, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.24073593318462372, + "learning_rate": 2.6045e-05, + "loss": 0.0488, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.2724616527557373, + "learning_rate": 2.6095e-05, + "loss": 0.0493, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.3012109398841858, + "learning_rate": 2.6145e-05, + "loss": 0.049, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23306235671043396, + "learning_rate": 2.6195e-05, + "loss": 0.0474, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.23879121243953705, + "learning_rate": 2.6245e-05, + "loss": 0.048, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20701922476291656, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0527, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.28734415769577026, + "learning_rate": 2.6345e-05, + "loss": 0.0469, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.26227426528930664, + "learning_rate": 2.6395e-05, + "loss": 0.0472, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.24196277558803558, + "learning_rate": 2.6445e-05, + "loss": 0.0462, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.3405964970588684, + "learning_rate": 2.6495e-05, + "loss": 0.0479, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.24121852219104767, + "learning_rate": 2.6545e-05, + "loss": 0.0444, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.284856379032135, + "learning_rate": 2.6595e-05, + "loss": 0.0492, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.2515311539173126, + "learning_rate": 2.6645e-05, + "loss": 0.0453, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.29297393560409546, + "learning_rate": 2.6695e-05, + "loss": 0.049, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.2528100311756134, + "learning_rate": 2.6745e-05, + "loss": 0.0477, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.29126009345054626, + "learning_rate": 2.6795e-05, + "loss": 0.0484, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.26489511132240295, + "learning_rate": 2.6845e-05, + "loss": 0.0476, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.33150023221969604, + "learning_rate": 2.6895e-05, + "loss": 0.0475, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.25904157757759094, + "learning_rate": 2.6945e-05, + "loss": 0.0495, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.24830900132656097, + "learning_rate": 2.6995e-05, + "loss": 0.0476, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.277997225522995, + "learning_rate": 2.7045e-05, + "loss": 0.0478, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.3100328743457794, + "learning_rate": 2.7095e-05, + "loss": 0.0501, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.2957077622413635, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0491, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.19402435421943665, + "learning_rate": 2.7195e-05, + "loss": 0.0426, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.20080482959747314, + "learning_rate": 2.7245e-05, + "loss": 0.0464, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.21148671209812164, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0457, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.2299162596464157, + "learning_rate": 2.7345e-05, + "loss": 0.0467, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2296467125415802, + "learning_rate": 2.7395e-05, + "loss": 0.0489, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.25099772214889526, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0487, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2503034472465515, + "learning_rate": 2.7495e-05, + "loss": 0.0533, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.24678738415241241, + "learning_rate": 2.7545e-05, + "loss": 0.0462, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.3281477987766266, + "learning_rate": 2.7595e-05, + "loss": 0.0492, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.32214412093162537, + "learning_rate": 2.7645e-05, + "loss": 0.0459, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.29244086146354675, + "learning_rate": 2.7695e-05, + "loss": 0.0495, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.2372455894947052, + "learning_rate": 2.7745e-05, + "loss": 0.0477, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.28411489725112915, + "learning_rate": 2.7795e-05, + "loss": 0.0514, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.36217001080513, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2913915514945984, + "learning_rate": 2.7895e-05, + "loss": 0.0535, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.3237503170967102, + "learning_rate": 2.7945e-05, + "loss": 0.0509, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.32016751170158386, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0526, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.25484171509742737, + "learning_rate": 2.8045e-05, + "loss": 0.0508, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.22641469538211823, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.23269280791282654, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.0484, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23346202075481415, + "learning_rate": 2.8195e-05, + "loss": 0.0472, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.25522616505622864, + "learning_rate": 2.8245e-05, + "loss": 0.0493, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.2478165179491043, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0468, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.25732970237731934, + "learning_rate": 2.8345e-05, + "loss": 0.0484, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.21047812700271606, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0455, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.252676397562027, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0496, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20048962533473969, + "learning_rate": 2.8495e-05, + "loss": 0.0434, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.1821785420179367, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0502, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.24106337130069733, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0471, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.413974404335022, + "learning_rate": 2.8645e-05, + "loss": 0.0542, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.29419073462486267, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0499, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.20694036781787872, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0453, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2390667200088501, + "learning_rate": 2.8795e-05, + "loss": 0.045, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2563699781894684, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.046, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.22678422927856445, + "learning_rate": 2.8895e-05, + "loss": 0.0456, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.26104822754859924, + "learning_rate": 2.8945e-05, + "loss": 0.0446, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.22175569832324982, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0478, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.20990405976772308, + "learning_rate": 2.9045e-05, + "loss": 0.0459, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.27277135848999023, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0477, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.22422750294208527, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.047, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.2740926742553711, + "learning_rate": 2.9195e-05, + "loss": 0.0444, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2977927327156067, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0477, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26431694626808167, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0488, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.22759316861629486, + "learning_rate": 2.9345e-05, + "loss": 0.0465, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.24671563506126404, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0501, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.24065586924552917, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0439, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.3334764838218689, + "learning_rate": 2.9495e-05, + "loss": 0.0475, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.23565605282783508, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.046, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.26102808117866516, + "learning_rate": 2.9595e-05, + "loss": 0.0494, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.34544840455055237, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.05, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.2476256936788559, + "learning_rate": 2.9695e-05, + "loss": 0.0477, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.2795628309249878, + "learning_rate": 2.9745e-05, + "loss": 0.0508, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.2756715714931488, + "learning_rate": 2.9795e-05, + "loss": 0.0484, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.20578329265117645, + "learning_rate": 2.9845e-05, + "loss": 0.0461, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.20354008674621582, + "learning_rate": 2.9895e-05, + "loss": 0.0431, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24710936844348907, + "learning_rate": 2.9945e-05, + "loss": 0.0459, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.32639119029045105, + "learning_rate": 2.9995e-05, + "loss": 0.045, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.27762553095817566, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0476, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.2903423607349396, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.3292105793952942, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0503, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.23503656685352325, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0446, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.24153585731983185, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.048, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.24157844483852386, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.046, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.23378467559814453, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0478, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.20753100514411926, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0527, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.2574090361595154, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0468, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.23341813683509827, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0445, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21378666162490845, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0487, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.2410328984260559, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0488, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.22005778551101685, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0478, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.3197212219238281, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.046, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.253745436668396, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.0437, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.1991635262966156, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0474, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.20773734152317047, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0439, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.3026197552680969, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0462, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.22027666866779327, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0443, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.2512618899345398, + "learning_rate": 2.99997744429684e-05, + "loss": 0.044, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.23001381754875183, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0479, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.20057332515716553, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0433, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.2831740975379944, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0462, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.18362601101398468, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0472, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.29741403460502625, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0443, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.20039376616477966, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0457, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.14999084174633026, + "learning_rate": 2.999958785135414e-05, + "loss": 0.043, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.18134453892707825, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0416, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.25010257959365845, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0494, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.31700536608695984, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0455, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.25080761313438416, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0436, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.2008715718984604, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0459, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.21441133320331573, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0492, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.23232029378414154, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0459, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.30173608660697937, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0505, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.21179363131523132, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0438, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.21905283629894257, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0439, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.17932383716106415, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0452, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.21132917702198029, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0454, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.22732721269130707, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.047, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.28214913606643677, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0446, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.23979637026786804, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0454, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.3095552921295166, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0449, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.2585015594959259, + "learning_rate": 2.999890232177341e-05, + "loss": 0.0473, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.18226519227027893, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0472, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2568792402744293, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.0443, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.27914026379585266, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0423, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.21452564001083374, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.0457, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.22575126588344574, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0458, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.24053922295570374, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.0453, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.21956735849380493, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.048, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2325354665517807, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0439, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.1692194789648056, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.044, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.25796884298324585, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0416, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.2719244956970215, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0445, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.23312833905220032, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0488, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.24458816647529602, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0473, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.24380528926849365, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0411, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.1836114227771759, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0433, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.21051307022571564, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0462, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.18093563616275787, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.045, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2444952130317688, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0485, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.19429947435855865, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0487, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.2294744998216629, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0479, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.2147439569234848, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0492, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.2777695059776306, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0419, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22944948077201843, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0458, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.2822294235229492, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0429, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.20201396942138672, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0435, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.1961182802915573, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.0441, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.1653209924697876, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0473, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.2082156240940094, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0508, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.199684277176857, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0466, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.20871016383171082, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0436, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.22781118750572205, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0504, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2652275562286377, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0467, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.20812813937664032, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0456, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2533996105194092, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0458, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.18139654397964478, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0441, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.27243369817733765, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0472, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.326709121465683, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0487, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.26698657870292664, + "learning_rate": 2.999617967180403e-05, + "loss": 0.05, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.4119804799556732, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0455, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.21175070106983185, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0462, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.22503416240215302, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.0468, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.254734605550766, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0471, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26713427901268005, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0439, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.2739362120628357, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0455, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.26221534609794617, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0451, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26752719283103943, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.23052464425563812, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0481, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.162479966878891, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0435, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.24716691672801971, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0459, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.14981329441070557, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0407, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.23326952755451202, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0459, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.2200203239917755, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.046, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.21501530706882477, + "learning_rate": 2.999465222590857e-05, + "loss": 0.0433, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.22950918972492218, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0454, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.2042679488658905, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.0444, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.16468611359596252, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0434, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.2180141657590866, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0447, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.1658746749162674, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0421, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.24921023845672607, + "learning_rate": 2.999396950511592e-05, + "loss": 0.0398, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0445, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.25189560651779175, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.0429, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.2614912688732147, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0427, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.2890806794166565, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0458, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.24215207993984222, + "learning_rate": 2.999336925698801e-05, + "loss": 0.04, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.1806204915046692, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0435, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.18884161114692688, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0422, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.2760990262031555, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0427, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.22660709917545319, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0459, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.24862384796142578, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.045, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.1911306381225586, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0433, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.2612020969390869, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0414, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.20945195853710175, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.043, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.2436099350452423, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0472, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.23788926005363464, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0455, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.20128341019153595, + "learning_rate": 2.999194851183094e-05, + "loss": 0.0456, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.24002164602279663, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0414, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.2653037905693054, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.042, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.15122903883457184, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0395, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.19299687445163727, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0452, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.2534617781639099, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0445, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.29057401418685913, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.0398, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.19744732975959778, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.041, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.25267109274864197, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0439, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2406993955373764, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0445, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.2561663091182709, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0465, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.2137032002210617, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0407, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.3002893030643463, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0433, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.32190272212028503, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0444, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.2076684534549713, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0432, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.31036099791526794, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0431, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.21386218070983887, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0442, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.19012001156806946, + "learning_rate": 2.998948184763285e-05, + "loss": 0.048, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.2310018688440323, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.0466, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.1984279453754425, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0443, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.24101053178310394, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0446, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.17791084945201874, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0403, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.2016495168209076, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0404, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.19435155391693115, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.21197110414505005, + "learning_rate": 2.998837053762875e-05, + "loss": 0.0398, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.22241441905498505, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0426, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.20825718343257904, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.0434, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.24533888697624207, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.046, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.2455352544784546, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0412, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.18937505781650543, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0425, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.22356615960597992, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0418, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.23797748982906342, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0457, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.23130296170711517, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0427, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.20548853278160095, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0439, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19323743879795074, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.0432, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.24339382350444794, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0412, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.18780124187469482, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0416, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.19353246688842773, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0461, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.21522872149944305, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0412, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.26847484707832336, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.3781932294368744, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0401, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.3219687342643738, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0423, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.1699237823486328, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0421, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.2174658179283142, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0398, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.2433294802904129, + "learning_rate": 2.998488804893431e-05, + "loss": 0.042, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.2116415947675705, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0405, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.22813065350055695, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0432, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.24252736568450928, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.048, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.2009168565273285, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0431, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.19379755854606628, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.23313777148723602, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0384, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.22056345641613007, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0459, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.3403800129890442, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0439, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.17649835348129272, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0419, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.20594412088394165, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0418, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.2901265323162079, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0405, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.22897225618362427, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0424, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.20961393415927887, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.0438, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.28585514426231384, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0477, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2239968478679657, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0398, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.1651894897222519, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0405, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.1731943041086197, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0407, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.1708967089653015, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0417, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.17396977543830872, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0359, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.21327704191207886, + "learning_rate": 2.998095036004704e-05, + "loss": 0.045, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.15585461258888245, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.0403, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.21774765849113464, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0403, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.1889292299747467, + "learning_rate": 2.99803204529095e-05, + "loss": 0.0408, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.1639491617679596, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0425, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.23745876550674438, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.043, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19481611251831055, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0453, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.21001650393009186, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0429, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.2114192098379135, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0419, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.2506379783153534, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0392, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.21439281105995178, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0422, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.24763159453868866, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0433, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.1719428300857544, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0422, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.25943970680236816, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0425, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22226977348327637, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.0448, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.2018195390701294, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0408, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.20869500935077667, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0443, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.23311494290828705, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.042, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.2501373589038849, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0382, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.2650911509990692, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0377, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.21386376023292542, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0405, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.27908873558044434, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0431, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.18731699883937836, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0402, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.2566622197628021, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0419, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.2418505847454071, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0406, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.199744313955307, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0424, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.20134931802749634, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.042, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.15607263147830963, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0453, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.22614715993404388, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0413, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.22810052335262299, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0433, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.16557516157627106, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0424, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.20254060626029968, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0394, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2068638950586319, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0408, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.27261367440223694, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.0433, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.23216570913791656, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0445, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.2546656131744385, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0381, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.21829853951931, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0472, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.17109332978725433, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0396, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.1746513843536377, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0413, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.22411789000034332, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.0373, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.21182285249233246, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.038, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.19430142641067505, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.0419, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.22196920216083527, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0395, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.20695635676383972, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0407, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.1721169799566269, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0396, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.23676829040050507, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0396, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.3019765317440033, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0443, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.23958736658096313, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0402, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.17959944903850555, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0434, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.1563226878643036, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0397, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.16746941208839417, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0402, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.26673364639282227, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0411, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.23634585738182068, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0427, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.19724313914775848, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0408, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.2072090059518814, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0389, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.18574072420597076, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0427, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.23390066623687744, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0411, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.2137259542942047, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0395, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.21458280086517334, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0422, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.1838904619216919, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.043, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.1834891140460968, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0389, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.2012249082326889, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0418, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.27410435676574707, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0473, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.21607685089111328, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0405, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.17525449395179749, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.043, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.23137639462947845, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0414, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.20958152413368225, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0404, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.21453802287578583, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.043, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.2522047758102417, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.04, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.23595386743545532, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0396, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.23375281691551208, + "learning_rate": 2.996358559911323e-05, + "loss": 0.042, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.21628527343273163, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0434, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.2463727742433548, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0391, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.19919127225875854, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.043, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.28369203209877014, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0463, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.30222833156585693, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0397, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.19382458925247192, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0402, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2075910121202469, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0392, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.2503066658973694, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0427, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.18662676215171814, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0397, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.2432677149772644, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0425, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.19900213181972504, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0422, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.12742096185684204, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0412, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.20504114031791687, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0404, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.19127227365970612, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0425, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.18699967861175537, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0419, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.20074906945228577, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0381, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.24741096794605255, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0389, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.22948721051216125, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0435, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.17772004008293152, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0412, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.228739395737648, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0406, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.18312685191631317, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0401, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.21340692043304443, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.0403, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.17110474407672882, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0388, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.19640640914440155, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.2386539727449417, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.043, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.20800922811031342, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0412, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1889747679233551, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0377, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.1801699548959732, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0402, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.20975525677204132, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0426, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.22626695036888123, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0404, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.2192756086587906, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0438, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.21307197213172913, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0447, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.28529244661331177, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0412, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.26367515325546265, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.044, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.29029133915901184, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.14019586145877838, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0427, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19176369905471802, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0408, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.19753597676753998, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.0395, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1900634765625, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0427, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.21698379516601562, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0452, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.21722441911697388, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0417, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.1760135143995285, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0405, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18030588328838348, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0425, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.20307210087776184, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0416, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.19713939726352692, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0401, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.18719582259655, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.19560874998569489, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.04, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1633717119693756, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0424, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.20183435082435608, + "learning_rate": 2.994811676932232e-05, + "loss": 0.0397, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.23784901201725006, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0438, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.17358708381652832, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0393, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18991878628730774, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0387, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.20455898344516754, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.042, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.23264537751674652, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0447, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2586922347545624, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0402, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.20447152853012085, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0406, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.1814388930797577, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0412, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2679784595966339, + "learning_rate": 2.994497917870469e-05, + "loss": 0.039, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.1818714141845703, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0423, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.2153957635164261, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0391, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.1858970820903778, + "learning_rate": 2.994391288470069e-05, + "loss": 0.0394, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.23248426616191864, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0439, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.17587599158287048, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0367, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.18720309436321259, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.039, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.17602671682834625, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0393, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.15901489555835724, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.0418, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.18207870423793793, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0401, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.22800953686237335, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0416, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.20031091570854187, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0417, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.24054989218711853, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0438, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.201461061835289, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0369, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.15697605907917023, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0397, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.16894802451133728, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0403, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2567543089389801, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0421, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.20297718048095703, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0396, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.17331846058368683, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.038, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.2214743047952652, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0405, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.2107590287923813, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0388, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.1768934726715088, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0364, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.16018562018871307, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0405, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.18402984738349915, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0425, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.21482129395008087, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0399, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.20554907619953156, + "learning_rate": 2.993578134823709e-05, + "loss": 0.0388, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.2401154637336731, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.0369, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.15072989463806152, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0359, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.1649719625711441, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0399, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.2386736422777176, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0403, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.2389005720615387, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.0416, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.14930105209350586, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.0372, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.18874907493591309, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0356, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.21466751396656036, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0391, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1697709560394287, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.24686385691165924, + "learning_rate": 2.993190369330068e-05, + "loss": 0.04, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.19733406603336334, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0417, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.3981468677520752, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0375, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.20926636457443237, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0384, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.18989330530166626, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0365, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17171184718608856, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0386, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.20474061369895935, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0413, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.19314105808734894, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0358, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.1794378012418747, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0409, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.17263059318065643, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0385, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.17863675951957703, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0355, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.20071984827518463, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0386, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.2025180608034134, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0409, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.21840184926986694, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0397, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.197452113032341, + "learning_rate": 2.992628447470613e-05, + "loss": 0.0379, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.15370187163352966, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0379, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.22221392393112183, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0396, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.18536874651908875, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.18744619190692902, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0378, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.21197383105754852, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0399, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.29221004247665405, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0355, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.2411249279975891, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.039, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.3187369108200073, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.036, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.18096214532852173, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0372, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.2743191719055176, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0401, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.1864418387413025, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0384, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.15518276393413544, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.039, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.18089881539344788, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0359, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.2218618243932724, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0427, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.17051172256469727, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0401, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.15884852409362793, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0378, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1561235636472702, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0383, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.19711007177829742, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0409, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.18595683574676514, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.043, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.18613699078559875, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0357, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.1390007734298706, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0356, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.25603482127189636, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0377, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.1902502328157425, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0381, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.18781402707099915, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0393, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.20504912734031677, + "learning_rate": 2.991569760600352e-05, + "loss": 0.042, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.17359918355941772, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0373, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.2029414027929306, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0412, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.23485742509365082, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0398, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1454097479581833, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0394, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.22343166172504425, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0386, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.17754614353179932, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0358, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.14276635646820068, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0405, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.2062550187110901, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.0429, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2093406766653061, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.038, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.14510402083396912, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0374, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.22726939618587494, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0373, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.14905233681201935, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0392, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.15828309953212738, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0361, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.990945968675917e-05, + "loss": 0.037, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.1867590844631195, + "learning_rate": 2.990900562792633e-05, + "loss": 0.038, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.18278446793556213, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0358, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.24628379940986633, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0393, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.2195759266614914, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0372, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.2397889941930771, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0352, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.1563243269920349, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0361, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.16278347373008728, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0366, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1776551604270935, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0369, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.19893717765808105, + "learning_rate": 2.990533239951709e-05, + "loss": 0.039, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.21403104066848755, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0378, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.19885864853858948, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0393, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.18425336480140686, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.0404, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.1930677592754364, + "learning_rate": 2.990346861711462e-05, + "loss": 0.0369, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18271560966968536, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0351, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.16518554091453552, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0367, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.17990152537822723, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0349, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.19625109434127808, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0368, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.1891077160835266, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0407, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.20162540674209595, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.17586444318294525, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0346, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.13146457076072693, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0338, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.16059468686580658, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0366, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.1724517047405243, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0395, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.1847251057624817, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0365, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.20104378461837769, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.037, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.2712373733520508, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0347, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.18014875054359436, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0388, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.17479945719242096, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0388, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.24176514148712158, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.0387, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.2453809529542923, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0381, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2105131894350052, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0369, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1763688176870346, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0384, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.16223029792308807, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0348, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.173785999417305, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.038, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.23269419372081757, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0393, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.19176329672336578, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0399, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.1559731364250183, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0375, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.15701691806316376, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0358, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.22872254252433777, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.0371, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15535064041614532, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0378, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.21613062918186188, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0382, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.1504523903131485, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0386, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.177171990275383, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2307550013065338, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0354, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.2215310037136078, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0371, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.23132716119289398, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0389, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.1937142014503479, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0373, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.23848143219947815, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.0355, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.23107734322547913, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0359, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1455046832561493, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0404, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.21890339255332947, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0421, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.1897345334291458, + "learning_rate": 2.988434806367412e-05, + "loss": 0.0357, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.2016463577747345, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0363, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.15850050747394562, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0391, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.18169008195400238, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0377, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.26095741987228394, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0391, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.20983633399009705, + "learning_rate": 2.988177232274603e-05, + "loss": 0.0355, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.21026623249053955, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.2104070633649826, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0343, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1801508367061615, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0369, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.22245049476623535, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.0358, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.19220919907093048, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.0354, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.2285826951265335, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0391, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.15382888913154602, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0374, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.13551367819309235, + "learning_rate": 2.987759237035982e-05, + "loss": 0.039, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.23362696170806885, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0367, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.15972478687763214, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0319, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.23858226835727692, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.0354, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.17988596856594086, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0391, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.21262454986572266, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0392, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.1443355828523636, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0384, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.24133095145225525, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0382, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.175334632396698, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.039, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.20340824127197266, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0376, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.16549751162528992, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0371, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.16527467966079712, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0373, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1418737769126892, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0378, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18754976987838745, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0355, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.13649867475032806, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0329, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.21408453583717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0348, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.20929333567619324, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0412, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.24083013832569122, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0378, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17924576997756958, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0369, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.22360439598560333, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0382, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.15963956713676453, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0361, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.19003787636756897, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0365, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2458433210849762, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0375, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.22048939764499664, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0382, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.19965334236621857, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0371, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.228081613779068, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0375, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.24077889323234558, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0347, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.17692111432552338, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0365, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.1996857225894928, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0358, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.24799935519695282, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0377, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.24300269782543182, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0362, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.2230309098958969, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0364, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.16640354692935944, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0346, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.14670157432556152, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.039, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18531887233257294, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0353, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.19566190242767334, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0391, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.1967019885778427, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0355, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.2161232978105545, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0375, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.17630350589752197, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0397, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18905608355998993, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0347, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.1890578716993332, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.036, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.15397298336029053, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0401, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.25086498260498047, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0362, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.20983757078647614, + "learning_rate": 2.985388676461079e-05, + "loss": 0.036, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.1554982215166092, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0378, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.1961488276720047, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0356, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.17265456914901733, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0357, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.1275556981563568, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0353, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.1667783409357071, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0354, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.27920690178871155, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0385, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.12451875954866409, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0352, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.15794622898101807, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0339, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.16285695135593414, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0369, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.18990163505077362, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0376, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.1867877095937729, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0363, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.15270273387432098, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0394, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.1934751272201538, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0382, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16943520307540894, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0369, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.20546840131282806, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0336, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.18490619957447052, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0415, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.16439616680145264, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0384, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.13623777031898499, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0379, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.18209043145179749, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0357, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.19922122359275818, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0365, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.15637168288230896, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0323, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1781233251094818, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0329, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.17547845840454102, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0344, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.1770639270544052, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0358, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.2677673101425171, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0356, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.16740617156028748, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0344, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.13076575100421906, + "learning_rate": 2.983793562696423e-05, + "loss": 0.038, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.17586535215377808, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0381, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.1623464673757553, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.037, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.16550037264823914, + "learning_rate": 2.983611255403156e-05, + "loss": 0.0373, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.20314498245716095, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.0366, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.19387134909629822, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0377, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.17531387507915497, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0381, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.21466659009456635, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.037, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1693795919418335, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0339, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.21929262578487396, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0332, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.2253660261631012, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0339, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.2084304839372635, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0363, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.2513981759548187, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0357, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.19585898518562317, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.0353, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.22262930870056152, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0357, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.19813011586666107, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0347, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.16626323759555817, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0346, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.271424800157547, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.036, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2283792346715927, + "learning_rate": 2.982684510938631e-05, + "loss": 0.04, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.15996390581130981, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0394, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.20269536972045898, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0387, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.20575514435768127, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0386, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.15178649127483368, + "learning_rate": 2.982433099619666e-05, + "loss": 0.037, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.23187878727912903, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0355, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.12988153100013733, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0369, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19367633759975433, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0367, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.15319304168224335, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0337, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.19311629235744476, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0383, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.14771778881549835, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0393, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.20381954312324524, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0414, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22292324900627136, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0358, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.13758333027362823, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0359, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.18837112188339233, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0339, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.1757460981607437, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0352, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.24256370961666107, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.0346, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.2417164295911789, + "learning_rate": 2.981603574229741e-05, + "loss": 0.037, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.21129392087459564, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0371, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16878797113895416, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0367, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.17526096105575562, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0353, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.1812993884086609, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.037, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.16877815127372742, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0353, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.20823006331920624, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0384, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1536571979522705, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0356, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.1477498710155487, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0346, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.12916848063468933, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0318, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.15418170392513275, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0369, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.1883322149515152, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0369, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.1569332480430603, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0374, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.21140694618225098, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0365, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.26859667897224426, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0395, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.2182101309299469, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0382, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.18896763026714325, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0372, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.17398814857006073, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0338, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.14043423533439636, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0361, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.14995983242988586, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0363, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.15542328357696533, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0361, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.16333894431591034, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0337, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.18489807844161987, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0352, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.1794673502445221, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.037, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.1882123351097107, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0414, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.16483181715011597, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.04, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.1727719008922577, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0361, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.1968923956155777, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0389, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1838955134153366, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.034, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.23556186258792877, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0369, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.2518751323223114, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0331, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18682284653186798, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0371, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.2049458622932434, + "learning_rate": 2.979480675356209e-05, + "loss": 0.04, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.11846950650215149, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.0347, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.21428292989730835, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0354, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.154558002948761, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0359, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1178995817899704, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0364, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.16926616430282593, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.0373, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.1721247136592865, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0346, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.1975203901529312, + "learning_rate": 2.979000948721388e-05, + "loss": 0.0351, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.17340779304504395, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0347, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.22424420714378357, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0384, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.1688208281993866, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.16542011499404907, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0343, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.15097306668758392, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0372, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.15530438721179962, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0345, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.16840992867946625, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.034, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.19959625601768494, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0375, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.15717248618602753, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0344, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.21473488211631775, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.04, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.18634898960590363, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0346, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.16032545268535614, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.0377, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.19153323769569397, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0408, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.16018801927566528, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0334, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.1498652696609497, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0349, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.1652575582265854, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0344, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.15636402368545532, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0331, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.22963644564151764, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0359, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.16931135952472687, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0389, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.17646029591560364, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.0322, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.19198739528656006, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0318, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.16826699674129486, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0389, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.1693607121706009, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.037, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.147409588098526, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0343, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.1984914094209671, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0329, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.18371893465518951, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0357, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.16279958188533783, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0377, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2122160941362381, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0367, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.20099543035030365, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0343, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.17249372601509094, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0325, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1106863021850586, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0336, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.2338913530111313, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0331, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.1545400768518448, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0321, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.17114387452602386, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0323, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.16218030452728271, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.037, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.14618389308452606, + "learning_rate": 2.976446886975759e-05, + "loss": 0.0306, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.17804935574531555, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0348, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1512002944946289, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0318, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.14877542853355408, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.037, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.1693819761276245, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0325, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.17713536322116852, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0356, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.1719064563512802, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.0358, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.19448526203632355, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0363, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.18047156929969788, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0326, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.20896470546722412, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0362, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.15999653935432434, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0318, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.17288383841514587, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0336, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.21483922004699707, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0367, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.19133612513542175, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0336, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.16010481119155884, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0377, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.1778886914253235, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0354, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.14997199177742004, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0326, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.16331784427165985, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0326, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.18247577548027039, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0367, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1838751882314682, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0338, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.17194764316082, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0365, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.1354181170463562, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0315, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.19473528861999512, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0345, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14786818623542786, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0354, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.1931883543729782, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0345, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.159915491938591, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0357, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14373569190502167, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0309, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.1468040943145752, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0339, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.1592605859041214, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0374, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.17073556780815125, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0365, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.13732466101646423, + "learning_rate": 2.974207486866042e-05, + "loss": 0.0331, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.18718381226062775, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0337, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.179783895611763, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0383, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.15795794129371643, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0368, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.13824236392974854, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.0343, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.2081076204776764, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0385, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.1407512128353119, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0338, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.1619846224784851, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.0325, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.184702068567276, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0353, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.18360818922519684, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0349, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.17683662474155426, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0332, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.17690211534500122, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0364, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.18250274658203125, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0336, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.12902002036571503, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0344, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.1598006635904312, + "learning_rate": 2.973127941557547e-05, + "loss": 0.0332, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.17691342532634735, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.037, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.2213180959224701, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0324, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.14762577414512634, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0321, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.14543993771076202, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.034, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.18970227241516113, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0377, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.15049827098846436, + "learning_rate": 2.972658565881519e-05, + "loss": 0.0341, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.1485714316368103, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.0361, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1893187314271927, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0354, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.2345002293586731, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0336, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.24119973182678223, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0325, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.18577414751052856, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0337, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.1941320151090622, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0355, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.17646151781082153, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0342, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.13801687955856323, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0351, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.15124225616455078, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0337, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19114945828914642, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.033, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.23505449295043945, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0357, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.15116260945796967, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.15656375885009766, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0322, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.18982963263988495, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0368, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.18300476670265198, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0379, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.16485247015953064, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0335, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.19399121403694153, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0353, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.22944431006908417, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0343, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.16755999624729156, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0351, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.15779320895671844, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0318, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.2113935351371765, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0343, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.14698673784732819, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0332, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19958600401878357, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0357, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20250289142131805, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0334, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.1955910474061966, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0381, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.13143742084503174, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0321, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.15840105712413788, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0333, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.20588421821594238, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0347, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.1881185919046402, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0345, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.16714425384998322, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.032, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.16526015102863312, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0316, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.1865321844816208, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0351, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.19202300906181335, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0325, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.15551044046878815, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0339, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.15684190392494202, + "learning_rate": 2.969840338753429e-05, + "loss": 0.034, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.16806578636169434, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0319, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.1554284542798996, + "learning_rate": 2.969675164571376e-05, + "loss": 0.0348, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1836974322795868, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0375, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.20003411173820496, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0364, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.1747797578573227, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.0328, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.2204974740743637, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0331, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.17183512449264526, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0381, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.14398148655891418, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.03, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.17428989708423615, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0329, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.13864420354366302, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0319, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.15052364766597748, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0357, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.21513856947422028, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0349, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19239531457424164, + "learning_rate": 2.968758727655112e-05, + "loss": 0.0377, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.22284342348575592, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0358, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.17206035554409027, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0349, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.1991468220949173, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0316, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.13542494177818298, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0329, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.14473386108875275, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0346, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.1972355991601944, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0348, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.17447993159294128, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0337, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.14419333636760712, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0355, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.19937965273857117, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0381, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.18988190591335297, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0331, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.1552465558052063, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0333, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.1769426017999649, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0344, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.17652767896652222, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0354, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1598699688911438, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0354, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.17556554079055786, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0372, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.18424467742443085, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.0372, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.1591036170721054, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.038, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.1973133534193039, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0332, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.139867901802063, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0336, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.17536301910877228, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0339, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16841425001621246, + "learning_rate": 2.966971695195351e-05, + "loss": 0.0319, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.24087557196617126, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.1419876366853714, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0341, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.17206361889839172, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.0315, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.14541317522525787, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0326, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.18411535024642944, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0382, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1769087016582489, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0355, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.16459472477436066, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0354, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.1770380288362503, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.034, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.16219645738601685, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0316, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.18393918871879578, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.0366, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.1436525285243988, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0318, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.18282611668109894, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.0376, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.16712592542171478, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0367, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.13408221304416656, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.0305, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.17664852738380432, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0312, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.13730520009994507, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0323, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1278112679719925, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0322, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.1742715984582901, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0339, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.15022823214530945, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0348, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.17816156148910522, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0348, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.18016202747821808, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0357, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.21158602833747864, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0331, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1747511923313141, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0301, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.16840165853500366, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0363, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.1507093459367752, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0352, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.2081453651189804, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.035, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.1459163874387741, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0331, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.17468400299549103, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0339, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.15824441611766815, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0309, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.19502659142017365, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0305, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.1502636969089508, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.032, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.13870884478092194, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0356, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1596919149160385, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0347, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.1830880492925644, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0359, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.13567551970481873, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0319, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.12907959520816803, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0327, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.17357082664966583, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.0308, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.20002517104148865, + "learning_rate": 2.963613172841619e-05, + "loss": 0.035, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.13852699100971222, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.0353, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.15445692837238312, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.11687827110290527, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0319, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.15454575419425964, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0366, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.1916663944721222, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0327, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.16530223190784454, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0309, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17331530153751373, + "learning_rate": 2.962976980219581e-05, + "loss": 0.032, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.17225238680839539, + "learning_rate": 2.962885651090167e-05, + "loss": 0.0323, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.16390499472618103, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0324, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.16459178924560547, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0315, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.1416798084974289, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0348, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.1546057015657425, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.0329, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15206973254680634, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0304, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.18791860342025757, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.03, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.17669418454170227, + "learning_rate": 2.962243236871436e-05, + "loss": 0.034, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.1335674524307251, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0323, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.17903365194797516, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0327, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.17771181464195251, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0341, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1532796174287796, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.0328, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.18777944147586823, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0343, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.2553282082080841, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0342, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.24115385115146637, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0347, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.18773671984672546, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0324, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.18168678879737854, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0355, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.14390279352664948, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0323, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.13260799646377563, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0336, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.16469483077526093, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0341, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.1618199497461319, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0305, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.15524879097938538, + "learning_rate": 2.960942086834339e-05, + "loss": 0.034, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.14689622819423676, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0352, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.17949312925338745, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0329, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.2376648485660553, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0326, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.1302678883075714, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0356, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.16316285729408264, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0318, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.14314667880535126, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0347, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.13162434101104736, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0335, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.13411720097064972, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0334, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.20853494107723236, + "learning_rate": 2.960094148810064e-05, + "loss": 0.036, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.15634609758853912, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0314, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.16109681129455566, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0323, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.15969422459602356, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0292, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.13126519322395325, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0338, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.13577662408351898, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.0323, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13226218521595, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0307, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.12283602356910706, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0305, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16819486021995544, + "learning_rate": 2.959332885815207e-05, + "loss": 0.0316, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.1443309634923935, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.031, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.17339679598808289, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0335, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.12366446852684021, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0319, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.18073806166648865, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0355, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.16230227053165436, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.033, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.15816937386989594, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0331, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.15911494195461273, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0296, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.19500406086444855, + "learning_rate": 2.958564529914256e-05, + "loss": 0.034, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.16553588211536407, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0348, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.1780945360660553, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0297, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16839291155338287, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0317, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.26763999462127686, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0335, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.1292608082294464, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0319, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19598962366580963, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0346, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.16073159873485565, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0317, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.2037622481584549, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.0324, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.11946277320384979, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0329, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.12799547612667084, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0329, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.23773977160453796, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0327, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.19241386651992798, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0371, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.15968935191631317, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0335, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.1712830811738968, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0302, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.16464142501354218, + "learning_rate": 2.957104758029206e-05, + "loss": 0.0346, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.2122460901737213, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0319, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.16571341454982758, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.033, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1610543429851532, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.0347, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14235250651836395, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0322, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.14999081194400787, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0299, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.1558496356010437, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.033, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.1448807269334793, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0307, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.20123368501663208, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0315, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.16191548109054565, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.031, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.1475522518157959, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.17617955803871155, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.0323, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.20633190870285034, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0287, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.13562145829200745, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0347, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.15800844132900238, + "learning_rate": 2.955719840324779e-05, + "loss": 0.034, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.1652759164571762, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0291, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.11819476634263992, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0309, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.17562547326087952, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0327, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.1934351772069931, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0342, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.1565367877483368, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.033, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.21048930287361145, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0357, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.16762706637382507, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0341, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.1267809122800827, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0322, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.18461033701896667, + "learning_rate": 2.954818092626432e-05, + "loss": 0.0357, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.13451576232910156, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0327, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.15478897094726562, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0311, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1461619883775711, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0314, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.1432616114616394, + "learning_rate": 2.954414442961676e-05, + "loss": 0.0339, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.14061163365840912, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0337, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.1950581669807434, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0315, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.13942912220954895, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.031, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.15901322662830353, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0324, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.19862298667430878, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.0322, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.15155215561389923, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0302, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16394968330860138, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0299, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.13225089013576508, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0305, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.18203529715538025, + "learning_rate": 2.953499770456317e-05, + "loss": 0.0319, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.15323573350906372, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0346, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.18822363018989563, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.038, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1510896533727646, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0336, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.12274178862571716, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.032, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.1287696361541748, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0341, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.16619601845741272, + "learning_rate": 2.952885021171824e-05, + "loss": 0.0321, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.12426833063364029, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0298, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.1751766800880432, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0341, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.13578996062278748, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0306, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.19475996494293213, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0314, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.13291390240192413, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0341, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.16193236410617828, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0325, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.16108311712741852, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0356, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.16233862936496735, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0303, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.14042457938194275, + "learning_rate": 2.951955450025455e-05, + "loss": 0.031, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.14922280609607697, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0336, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.1389254927635193, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0324, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.21042059361934662, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.0331, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.18529878556728363, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0311, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1345708668231964, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0316, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.14794935286045074, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0367, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.18609392642974854, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0326, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.1762595772743225, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0312, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.15549533069133759, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0342, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.1910371482372284, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0316, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.12789981067180634, + "learning_rate": 2.950807178791257e-05, + "loss": 0.0286, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.11457879841327667, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0303, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.12534750998020172, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0313, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.17089471220970154, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0318, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.16774669289588928, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0326, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.18693485856056213, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0356, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.20419354736804962, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.0306, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15851624310016632, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0328, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1284184604883194, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.1148369237780571, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0291, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.15302640199661255, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.037, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.16082480549812317, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.031, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.16006684303283691, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0346, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14866025745868683, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0323, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.15490789711475372, + "learning_rate": 2.949326462724779e-05, + "loss": 0.033, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.18960629403591156, + "learning_rate": 2.949219871528611e-05, + "loss": 0.0302, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.16329076886177063, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0338, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1700190156698227, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0326, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.14541998505592346, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0332, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.13809162378311157, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.16794940829277039, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0345, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1897762268781662, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0304, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2064688503742218, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0332, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.14193041622638702, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0332, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.15571711957454681, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0339, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.14264710247516632, + "learning_rate": 2.948147907679037e-05, + "loss": 0.036, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.1411963850259781, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0344, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.18891720473766327, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0308, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.15993206202983856, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0339, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.15609434247016907, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.0315, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.19272638857364655, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0325, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.2836095690727234, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0323, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16382543742656708, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0335, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.18899215757846832, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0301, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.17764721810817719, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0322, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.1808568239212036, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0304, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.166172593832016, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.0298, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.18262767791748047, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0305, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.17854458093643188, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.033, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.16440987586975098, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.033, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.19478020071983337, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0288, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.2605016231536865, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0318, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1850014477968216, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0313, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.1339086890220642, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.17828164994716644, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0315, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.1235261932015419, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0328, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.13704872131347656, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0322, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1296597719192505, + "learning_rate": 2.945750886818645e-05, + "loss": 0.0315, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.16906750202178955, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0347, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14643864333629608, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.031, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.16102853417396545, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0315, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.17575663328170776, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0312, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1488381028175354, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0318, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.14629733562469482, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0333, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.17051763832569122, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.034, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.14636048674583435, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0315, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.19957765936851501, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.032, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.15027561783790588, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0342, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1580146700143814, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.032, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1452188491821289, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0302, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.1924576759338379, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0323, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.199595108628273, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0326, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.1557701975107193, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0303, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.21232250332832336, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0324, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.13840343058109283, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0321, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.14219941198825836, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0314, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1747855842113495, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.0342, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1730964183807373, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0334, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.12938249111175537, + "learning_rate": 2.94341324778745e-05, + "loss": 0.029, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1782408058643341, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0299, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.15676888823509216, + "learning_rate": 2.943188093053675e-05, + "loss": 0.034, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.12131186574697495, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0277, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.19141671061515808, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0307, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.2431664764881134, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0326, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15145368874073029, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0329, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.19781085848808289, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0332, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.19477561116218567, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0317, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.1818031370639801, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0352, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.13725785911083221, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0272, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.1874770075082779, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0322, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.18701452016830444, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0304, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.15173715353012085, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.032, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.1315767467021942, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0331, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.16627946496009827, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0299, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.2122701108455658, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0308, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.12093837559223175, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0303, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.2072226107120514, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0331, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.14395366609096527, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0323, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.18454021215438843, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.0349, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1647966057062149, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0342, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1907777190208435, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0347, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.14438189566135406, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0295, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.16802345216274261, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0283, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.1488419622182846, + "learning_rate": 2.940567320758981e-05, + "loss": 0.033, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14881044626235962, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.0318, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.18762606382369995, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0322, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.15426765382289886, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0314, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.17882433533668518, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0361, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.16833186149597168, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0319, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.16899487376213074, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0283, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.15301142632961273, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0325, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.13270439207553864, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0305, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.16603541374206543, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0347, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.21959784626960754, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.0299, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.24568438529968262, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0339, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.14889934659004211, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0284, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.20354624092578888, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0309, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.18124797940254211, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0352, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.13871392607688904, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0311, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.18997471034526825, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0299, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.09903991222381592, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0304, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.13336266577243805, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0306, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.1564595103263855, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0295, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.17495550215244293, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.17048844695091248, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0266, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.1257491409778595, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0299, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.15244479477405548, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0341, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.15737923979759216, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.03, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.13226579129695892, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0268, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.21549297869205475, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0323, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.2102872133255005, + "learning_rate": 2.937416924082395e-05, + "loss": 0.029, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.14369380474090576, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0306, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.1390914022922516, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0273, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.13536471128463745, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0299, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.13321688771247864, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0298, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13358351588249207, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0309, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.1732599437236786, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0319, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.17455634474754333, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0314, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.1714484691619873, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0345, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.18120428919792175, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0293, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.21681863069534302, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0335, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.12099924683570862, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0299, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.13233523070812225, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0372, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.16708864271640778, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0305, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.2122814804315567, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0319, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14159013330936432, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0343, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.13396364450454712, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0308, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.17630934715270996, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0349, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.16417403519153595, + "learning_rate": 2.935272438143983e-05, + "loss": 0.0316, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.16593796014785767, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0291, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.13613900542259216, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0299, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.15980054438114166, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0319, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1578991562128067, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0288, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.1628173291683197, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0317, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.15937024354934692, + "learning_rate": 2.934549758921277e-05, + "loss": 0.031, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.15258124470710754, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0323, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.16142596304416656, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0289, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.15643659234046936, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0295, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.17128989100456238, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0313, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.1871197372674942, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0314, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.1448175013065338, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0318, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.21338209509849548, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0334, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15262271463871002, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.0293, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.14302000403404236, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0307, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.13756762444972992, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0288, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.1390128880739212, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0286, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.16088609397411346, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0299, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.11852050572633743, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0284, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.17093200981616974, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0331, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.2325763702392578, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0269, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.19948336482048035, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0314, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.16918544471263885, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0311, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.14563684165477753, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0301, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.19645430147647858, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0312, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.18950732052326202, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0292, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.11659497767686844, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0307, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1314595341682434, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0326, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1457567662000656, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.16525784134864807, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0322, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.15977199375629425, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0334, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.17081087827682495, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0308, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.20844970643520355, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.029, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.22166086733341217, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1766587346792221, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0298, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1307239532470703, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0298, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.17360512912273407, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0289, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.17264464497566223, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0317, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.15076857805252075, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0308, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14443998038768768, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0291, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.13115015625953674, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0329, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1444655954837799, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0294, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.16740760207176208, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0318, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.14201150834560394, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0314, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.14225733280181885, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0297, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.16758595407009125, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.0307, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.16751281917095184, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.03, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.1855877935886383, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0333, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.17420977354049683, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0277, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.1568649560213089, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0312, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.15874658524990082, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0297, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.13727585971355438, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0334, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.14999760687351227, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0328, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.16789689660072327, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0305, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.14890174567699432, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0312, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.18001733720302582, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0322, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.1718987226486206, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.03, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.18727312982082367, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0297, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.1528918594121933, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0296, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.13372628390789032, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.0288, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.14980314671993256, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0319, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.11368083208799362, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0312, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.14968374371528625, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0311, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.15593667328357697, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0287, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.1762111783027649, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0296, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.15493011474609375, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.0272, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.14142747223377228, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.03, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.1307222545146942, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0294, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.1598598062992096, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.03, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.17966923117637634, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0337, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15533174574375153, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0276, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.1223452091217041, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0294, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.12733329832553864, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0289, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.15903618931770325, + "learning_rate": 2.926085774098786e-05, + "loss": 0.029, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1539735347032547, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0272, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.20487286150455475, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0312, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.19920656085014343, + "learning_rate": 2.925700806726565e-05, + "loss": 0.0315, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.15179695188999176, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0305, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.16460831463336945, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0315, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.21031928062438965, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0283, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.13835182785987854, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0287, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.1597253829240799, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0336, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.11488606780767441, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0328, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.18707934021949768, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.0328, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.18874959647655487, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0317, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.1182878389954567, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0316, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.11695742607116699, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0308, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.1597018986940384, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0324, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.17729975283145905, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0316, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.18250048160552979, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0309, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.20568659901618958, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0293, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.12907297909259796, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0281, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.1809365600347519, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0333, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.17379151284694672, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0327, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.16082972288131714, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0308, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.17415443062782288, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.0299, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.1659587323665619, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0272, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.1538572907447815, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.0309, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.16303570568561554, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0322, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1444631963968277, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0306, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.24458004534244537, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0285, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.15384158492088318, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0277, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1425933837890625, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0303, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.12485234439373016, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.0279, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.14287105202674866, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0297, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.1621903032064438, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0268, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.145540252327919, + "learning_rate": 2.921797581830774e-05, + "loss": 0.03, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.1410767287015915, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0309, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.10953503102064133, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0299, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.15029482543468475, + "learning_rate": 2.92140191009304e-05, + "loss": 0.031, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.13262124359607697, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0299, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1503380835056305, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0329, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.18932904303073883, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.028, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.1473148614168167, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0315, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.18515919148921967, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0309, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.16122235357761383, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0279, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.1684419959783554, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0309, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.21400189399719238, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0304, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.12552157044410706, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0286, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.15383073687553406, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.029, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.1269458532333374, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0289, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.1839684545993805, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0316, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.13441112637519836, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0281, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.13045959174633026, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0296, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.13933542370796204, + "learning_rate": 2.919408984114327e-05, + "loss": 0.029, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.15143601596355438, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0296, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.16731251776218414, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.032, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.13245129585266113, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.0268, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.17244738340377808, + "learning_rate": 2.918873439576597e-05, + "loss": 0.028, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.1709858626127243, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0293, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18727441132068634, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0302, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1482084095478058, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0297, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.1343366652727127, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0303, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.1456393450498581, + "learning_rate": 2.918201584514458e-05, + "loss": 0.0299, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1569104641675949, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0281, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.19711613655090332, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0298, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.14027468860149384, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.031, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.16852174699306488, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0289, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.18446287512779236, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0308, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.16207733750343323, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0283, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.19747264683246613, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0301, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.15206509828567505, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0295, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.18873688578605652, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0286, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.144051656126976, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0335, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.14620175957679749, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.0284, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.19520114362239838, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0304, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.16127106547355652, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0333, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.11467311531305313, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0303, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11777719110250473, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0311, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.12151405215263367, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0278, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1852089911699295, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.0297, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.19555963575839996, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0321, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.1584533303976059, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.0292, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.1764054298400879, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0306, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.16257385909557343, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0298, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.13896049559116364, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.0296, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.16509538888931274, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0321, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.14912301301956177, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0285, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.18058481812477112, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.0302, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.18734610080718994, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0289, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.19707795977592468, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.0327, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.16215947270393372, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0302, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1710345447063446, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0332, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.19638675451278687, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0316, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.17807847261428833, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.0285, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.13119569420814514, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0298, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.13414712250232697, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0301, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.1941734254360199, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0327, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.19624412059783936, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.0291, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.17615875601768494, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0293, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.19011878967285156, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0316, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.15080299973487854, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0263, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.1676572859287262, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0318, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.14975300431251526, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0263, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17691564559936523, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.17775340378284454, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0298, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.14438258111476898, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.0275, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.14729741215705872, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0289, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.20924119651317596, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0297, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.15471352636814117, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0289, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.1571032702922821, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0307, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15048687160015106, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0301, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.15063059329986572, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0279, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15528838336467743, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0299, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.1756942719221115, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0291, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.1491880565881729, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0296, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.1562451720237732, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0302, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.12366753816604614, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0257, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.15308785438537598, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.16614912450313568, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0301, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.12428635358810425, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0303, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.135500967502594, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0283, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.16944879293441772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.0279, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.174363911151886, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0295, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.16406473517417908, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0284, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.2144644856452942, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0277, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.22006124258041382, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0264, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.15261462330818176, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0298, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.15205760300159454, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.14101643860340118, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0305, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16015839576721191, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.17772945761680603, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0312, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16818812489509583, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0283, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1448637843132019, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0325, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.14702047407627106, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0303, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.15547925233840942, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0299, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.18527188897132874, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.029, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.1533391922712326, + "learning_rate": 2.907944011202334e-05, + "loss": 0.029, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.15355326235294342, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0286, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.1503538340330124, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0312, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.12572433054447174, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0277, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.13614071905612946, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0303, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.13363786041736603, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0292, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.13318102061748505, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.0289, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.14687903225421906, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0301, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.13144652545452118, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0271, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.16351011395454407, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.0292, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.19256997108459473, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0268, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13006120920181274, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.0285, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.12932874262332916, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.11981740593910217, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.14302268624305725, + "learning_rate": 2.905937392537971e-05, + "loss": 0.029, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.1446913331747055, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0289, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.1500432789325714, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0264, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.15604284405708313, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0283, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.1795264482498169, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0289, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.14356553554534912, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0304, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.1323058009147644, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0302, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.2098175585269928, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0297, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.1424631029367447, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0295, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.1956447958946228, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0294, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.11993541568517685, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0273, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.18414588272571564, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0316, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.17320625483989716, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0319, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.17650462687015533, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0304, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.18740026652812958, + "learning_rate": 2.903909846652783e-05, + "loss": 0.034, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.12898528575897217, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0309, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.11721221357584, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0297, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.1659565567970276, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0294, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.1469241976737976, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0298, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14588896930217743, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0332, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.16243143379688263, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0305, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.15857404470443726, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0318, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.14992372691631317, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0305, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.13760966062545776, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0299, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.14302772283554077, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0313, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.18560320138931274, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0262, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.13117150962352753, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0286, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.16370882093906403, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0298, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.18689852952957153, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.031, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.16412676870822906, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0289, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.20153313875198364, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0297, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.15255704522132874, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0316, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.2036312222480774, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0306, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.16087041795253754, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.0315, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.1514595001935959, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0295, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.13961558043956757, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0259, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1996241807937622, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0336, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.15389376878738403, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0278, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.12345903366804123, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.032, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.16360828280448914, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0292, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.19773486256599426, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.032, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.11931311339139938, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0268, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.1378420740365982, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0296, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.09076640754938126, + "learning_rate": 2.899643488777395e-05, + "loss": 0.026, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.1116260215640068, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0266, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.18670536577701569, + "learning_rate": 2.899345958962103e-05, + "loss": 0.0314, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.15351636707782745, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0279, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.15310993790626526, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0298, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.15458840131759644, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0329, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.12034284323453903, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0266, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.14272604882717133, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0296, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.16335424780845642, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0266, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.16105882823467255, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0355, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.1455371230840683, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0261, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.21555180847644806, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0301, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.15278705954551697, + "learning_rate": 2.897851935432475e-05, + "loss": 0.0292, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.14099596440792084, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0287, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.10848120599985123, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.028, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.1380046308040619, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0273, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.14006251096725464, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0328, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.16518674790859222, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0286, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.13295377790927887, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.031, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.1623251736164093, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0281, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.16029980778694153, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.03, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.15575948357582092, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0294, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.157670795917511, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0294, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1470523178577423, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0281, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.12192463129758835, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0248, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.1240374892950058, + "learning_rate": 2.895893835994484e-05, + "loss": 0.028, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1385754942893982, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0287, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.140727236866951, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0281, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.15758906304836273, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0294, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.19664499163627625, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0277, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.12567110359668732, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0279, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.13595536351203918, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0313, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.14262230694293976, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0317, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1474338322877884, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0282, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.18601401150226593, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0274, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.1392786204814911, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.0269, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.12022347748279572, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0262, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.17340584099292755, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0277, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.1438349187374115, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0273, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.1408676654100418, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0277, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.14853249490261078, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0291, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.20401965081691742, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.031, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.18385861814022064, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.12695786356925964, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0267, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.14115597307682037, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0291, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.15790492296218872, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.0294, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.15104074776172638, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0278, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13686735928058624, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0276, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.12029985338449478, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0305, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1351509988307953, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0293, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.10609307885169983, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.0285, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15056312084197998, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0292, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.17062142491340637, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.0281, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.15565407276153564, + "learning_rate": 2.891615575134314e-05, + "loss": 0.0298, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.136429563164711, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0282, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.1649797409772873, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0267, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.19815988838672638, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.032, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15660062432289124, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.03, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.1840171217918396, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0342, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.14063017070293427, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0282, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.14935986697673798, + "learning_rate": 2.890533053632388e-05, + "loss": 0.029, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.14691074192523956, + "learning_rate": 2.89037798520224e-05, + "loss": 0.033, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.14992092549800873, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0262, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.16885656118392944, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0301, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1223178282380104, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0275, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.1200852021574974, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0277, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.1251501888036728, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0252, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.16905754804611206, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0277, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.13312722742557526, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0296, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.13198228180408478, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0269, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.14326612651348114, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0253, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.14861920475959778, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0262, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.19368910789489746, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0295, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.143910750746727, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0282, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.12314137071371078, + "learning_rate": 2.888352491214935e-05, + "loss": 0.027, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.16095006465911865, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0272, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.1414606273174286, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.029, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.15631091594696045, + "learning_rate": 2.887882538107974e-05, + "loss": 0.0291, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16476406157016754, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0283, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.2236901968717575, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0286, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.1318763792514801, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0274, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.15661245584487915, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.029, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.13406990468502045, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0287, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16861270368099213, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0272, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.18082283437252045, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0285, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.17517460882663727, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0277, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.17011325061321259, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0325, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14378181099891663, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0306, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1766553819179535, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0304, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.1666528880596161, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0294, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.15178363025188446, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0273, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.14275510609149933, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0291, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.1568918377161026, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0267, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.1370164304971695, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0268, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.10388749092817307, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0289, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11448585987091064, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0302, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.15148931741714478, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0292, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.14388935267925262, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0257, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.14465516805648804, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0272, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.13414210081100464, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0273, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12774589657783508, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.028, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.24874050915241241, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0282, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.1624344289302826, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0294, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.12817884981632233, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0324, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1509743630886078, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0269, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.31345340609550476, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0287, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.1523321270942688, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0259, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.1633133739233017, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0269, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.14278745651245117, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.0253, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.18361203372478485, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0282, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.10244551301002502, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0268, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.125647634267807, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0264, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.12238023430109024, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0264, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.16535502672195435, + "learning_rate": 2.882169202177678e-05, + "loss": 0.0329, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.19624854624271393, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0303, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.18285983800888062, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0266, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.13460052013397217, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.0293, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.1427820920944214, + "learning_rate": 2.881525979398615e-05, + "loss": 0.0244, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.14855900406837463, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0276, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.1433439403772354, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0299, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.1452850103378296, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0271, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.14403793215751648, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0264, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.13841332495212555, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0277, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.1426801085472107, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0296, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.14492706954479218, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0272, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.1315297782421112, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.0289, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.1513436734676361, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0284, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.15784604847431183, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.0273, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.12729156017303467, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0297, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.20208381116390228, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0284, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.1486770659685135, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0271, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.1463368982076645, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.0292, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.12362723797559738, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0252, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.18984590470790863, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0292, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.12962038815021515, + "learning_rate": 2.878773564250086e-05, + "loss": 0.0301, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.132072314620018, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0298, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.1482372134923935, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0267, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.15687556564807892, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0308, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.11656040698289871, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.028, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.16482318937778473, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0256, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.13332504034042358, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.027, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.16453887522220612, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0253, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.16902299225330353, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0283, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.12483953684568405, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0276, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.14033488929271698, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0277, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.18922871351242065, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0258, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.15026360750198364, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0298, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17492659389972687, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0275, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.14793051779270172, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0301, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.17575892806053162, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.027, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.1661873608827591, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0275, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.14529843628406525, + "learning_rate": 2.875990888327069e-05, + "loss": 0.0277, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.11084084957838058, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0278, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1433780938386917, + "learning_rate": 2.875661528053842e-05, + "loss": 0.026, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.13195206224918365, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0298, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.14103750884532928, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0309, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.17067542672157288, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0296, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.14078323543071747, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0288, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.17906904220581055, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0268, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.163010373711586, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0285, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.17636623978614807, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0276, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.18726587295532227, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0297, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.18583180010318756, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.16773444414138794, + "learning_rate": 2.874008460348417e-05, + "loss": 0.0289, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.14579996466636658, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.12655268609523773, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0246, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.15124735236167908, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0297, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.16273140907287598, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0295, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14014655351638794, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0262, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12657561898231506, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0292, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.13813915848731995, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.0287, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.12874144315719604, + "learning_rate": 2.872678492608539e-05, + "loss": 0.026, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.12086840718984604, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0245, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.15290603041648865, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0297, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.13261260092258453, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0263, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.17371073365211487, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0268, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.17300288379192352, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0264, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.16441872715950012, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0279, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.17883789539337158, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.03, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.14519335329532623, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0266, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.18063998222351074, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0275, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.19818824529647827, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0256, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1695982664823532, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0274, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.16905854642391205, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0277, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.1852719783782959, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.03, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.14601513743400574, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0261, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.11720956116914749, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0279, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.14177089929580688, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0268, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.14933669567108154, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0292, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15183576941490173, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0285, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.1255270391702652, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.025, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.1467735320329666, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0258, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.13636574149131775, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0278, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.12642863392829895, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0254, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.09835807979106903, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0284, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.11784205585718155, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0285, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.14453290402889252, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0301, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.19569514691829681, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0264, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.13860125839710236, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.026, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.16651827096939087, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0283, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.13682647049427032, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0309, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.14632774889469147, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0275, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.1401532143354416, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0261, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.12643243372440338, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0268, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1152961477637291, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0259, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.14027607440948486, + "learning_rate": 2.866951776928e-05, + "loss": 0.0306, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.12379784137010574, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0272, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.15595580637454987, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0288, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.12823393940925598, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0263, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1481468379497528, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0283, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.12768998742103577, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0262, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.15771906077861786, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0267, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.18978171050548553, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0295, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.1386374831199646, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0277, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.1583528220653534, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0273, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.09999477118253708, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0253, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.14432361721992493, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0289, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.19001086056232452, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.0295, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.13931159675121307, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0257, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.1621151715517044, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0269, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.15239235758781433, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0275, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12619581818580627, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0273, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.14698320627212524, + "learning_rate": 2.864043385528954e-05, + "loss": 0.0275, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.14121954143047333, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0271, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.16203583776950836, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0265, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.15562862157821655, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0267, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.16438758373260498, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0285, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.13769859075546265, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0274, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.14908984303474426, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0256, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.14057780802249908, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.0261, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.16056638956069946, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.0267, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.12299710512161255, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0257, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.12815117835998535, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0284, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.14666494727134705, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0303, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.15829558670520782, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0243, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1499258577823639, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.029, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.16584458947181702, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0272, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1765419989824295, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0278, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.17066679894924164, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0281, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13183729350566864, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0255, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.12179457396268845, + "learning_rate": 2.86093128283797e-05, + "loss": 0.028, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.10134928673505783, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0275, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16393089294433594, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0278, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1360260546207428, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0297, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.16158054769039154, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0249, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.12037541717290878, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.0286, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.16677190363407135, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0277, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1442420333623886, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0268, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.1756538599729538, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0282, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.11580783128738403, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0282, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.11671322584152222, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0279, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.15163642168045044, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0297, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.1390339583158493, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0282, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.11784064769744873, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0275, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.15270282328128815, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0282, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.09030280262231827, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0281, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.15108083188533783, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0274, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.1403200924396515, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0265, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.14890284836292267, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0303, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.11579708009958267, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0283, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.14590834081172943, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0268, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.1513289511203766, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0264, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.13494084775447845, + "learning_rate": 2.857082133984191e-05, + "loss": 0.026, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.11738834530115128, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.029, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.1497882753610611, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0271, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.17350555956363678, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0277, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.16718043386936188, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0269, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.12658995389938354, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0267, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15880689024925232, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0256, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.1480095237493515, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0263, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.13779640197753906, + "learning_rate": 2.85567006866004e-05, + "loss": 0.0282, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1088612824678421, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0245, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.17420272529125214, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0267, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.12365500628948212, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.025, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.13261641561985016, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0256, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.10485266894102097, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0253, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14070947468280792, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0284, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.1179763451218605, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0241, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13687312602996826, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0284, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.17160922288894653, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0263, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.17487400770187378, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0275, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.11352817714214325, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1871306300163269, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0277, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.13434743881225586, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0278, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.16361933946609497, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0257, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.19457586109638214, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0277, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.16905367374420166, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0323, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.14845560491085052, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0248, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.12934044003486633, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.026, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.11046823114156723, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0267, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.14174290001392365, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0255, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.15194296836853027, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.0257, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.1532754749059677, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0269, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19864243268966675, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0288, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.16765300929546356, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0271, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.15602844953536987, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0288, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.1571321040391922, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0264, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1625281423330307, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0268, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.12808986008167267, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0307, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.12471310794353485, + "learning_rate": 2.850496147627652e-05, + "loss": 0.027, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.1521507054567337, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0289, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1389119178056717, + "learning_rate": 2.850136142292186e-05, + "loss": 0.024, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1338133066892624, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13214652240276337, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.0274, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.17002162337303162, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0277, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.1433296948671341, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0267, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.19547788798809052, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0287, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.1916888952255249, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0296, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.17188303172588348, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0262, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.14335697889328003, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.0269, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14298376441001892, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0278, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.12892583012580872, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0278, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.16246432065963745, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0255, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.1237327978014946, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0252, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.12249717861413956, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0238, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.16828545928001404, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0262, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.12834671139717102, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0287, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.11583935469388962, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.029, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.10479022562503815, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0244, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.18407097458839417, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0276, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.12898766994476318, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0277, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.13034385442733765, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0241, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.13197654485702515, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0269, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.14327986538410187, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0281, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.15228267014026642, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0269, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.11858481913805008, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0261, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.16075590252876282, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0284, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.1527116447687149, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0283, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.1280883252620697, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0294, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1348433941602707, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0251, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.15039700269699097, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0272, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.16682399809360504, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.0277, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.11187390983104706, + "learning_rate": 2.844503681416359e-05, + "loss": 0.0298, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.1675758808851242, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0267, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.16022244095802307, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0301, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.16192372143268585, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0277, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.13092738389968872, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.12416383624076843, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.028, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.12514744699001312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0252, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.1705007702112198, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.0274, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.14251261949539185, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0264, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.147093266248703, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0279, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.12747326493263245, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0265, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.14222867786884308, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0276, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.12574070692062378, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0262, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.12812504172325134, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0225, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.13585571944713593, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0267, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.12168630957603455, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0236, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.12048719823360443, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0225, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.15165774524211884, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0267, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.15395723283290863, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.026, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.14715734124183655, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0259, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.11370360106229782, + "learning_rate": 2.840817735753433e-05, + "loss": 0.0284, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1264869123697281, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.028, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.13726982474327087, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0268, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1450515240430832, + "learning_rate": 2.840261328154438e-05, + "loss": 0.0251, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.1328781247138977, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.025, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.15474216639995575, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.0292, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.1338219791650772, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0279, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.16368664801120758, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.027, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.18125605583190918, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.0252, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.127846822142601, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0252, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.17743080854415894, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0275, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.16253551840782166, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0287, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.14212867617607117, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0274, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.1380675882101059, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0263, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.12773451209068298, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.0257, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.1556098759174347, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0264, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.16690294444561005, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0296, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.11710768938064575, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0252, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.17441801726818085, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0318, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.10790088772773743, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0251, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.16537238657474518, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0266, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.1414048671722412, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0276, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.1096765398979187, + "learning_rate": 2.836716156259365e-05, + "loss": 0.0258, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.14450953900814056, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0278, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15733887255191803, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0278, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.15253902971744537, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0275, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.1410975307226181, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0249, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1242721900343895, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0251, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.15358713269233704, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0287, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.15856830775737762, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.029, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.1777517944574356, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0254, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.10823221504688263, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0262, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.144780695438385, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0286, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12900282442569733, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0243, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.15370796620845795, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0238, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.15801627933979034, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0259, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.19058606028556824, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0257, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14663439989089966, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0281, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.10609344393014908, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.0261, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.11403398960828781, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0291, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.11631747335195541, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0243, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.1325336992740631, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0266, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.11245301365852356, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.10258981585502625, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0252, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.15847361087799072, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0258, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.15267148613929749, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0246, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.15297843515872955, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0248, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.1767774075269699, + "learning_rate": 2.831995639121568e-05, + "loss": 0.0292, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.12696890532970428, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.025, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.17774799466133118, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0267, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.15603452920913696, + "learning_rate": 2.83142492407012e-05, + "loss": 0.0285, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.12731599807739258, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0272, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.13816066086292267, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0265, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.14216956496238708, + "learning_rate": 2.830853299002291e-05, + "loss": 0.026, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1427476406097412, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0263, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.14667609333992004, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0282, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.11785806715488434, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0276, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.125101238489151, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0255, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.10735481232404709, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0248, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.14256443083286285, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0286, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12700040638446808, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0275, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.11069332808256149, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0283, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.1438131332397461, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.0269, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.1901932805776596, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0274, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.17823027074337006, + "learning_rate": 2.828557706390495e-05, + "loss": 0.029, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.13579529523849487, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0267, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.1431250274181366, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0279, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.11313948035240173, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0274, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1669740527868271, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0288, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.12642791867256165, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0267, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.15632158517837524, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0256, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.1404937356710434, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0263, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.13466882705688477, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0249, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.13014933466911316, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0255, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.18688003718852997, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0267, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.14594615995883942, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0253, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.1708635836839676, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0244, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.16193392872810364, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0265, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.13249778747558594, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0286, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.13875705003738403, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0262, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.19137892127037048, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0246, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.15596376359462738, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0248, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.1577422320842743, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0247, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.14322742819786072, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0263, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.13651713728904724, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0261, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11288725584745407, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0252, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.09636299312114716, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0267, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.11729153245687485, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0275, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.14827747642993927, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0265, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.1446153074502945, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0257, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.1329014152288437, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0247, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.12726181745529175, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0259, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.13508957624435425, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0258, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.14538758993148804, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0264, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.10585593432188034, + "learning_rate": 2.822755216255228e-05, + "loss": 0.0276, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.17435809969902039, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0281, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.1504661589860916, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0259, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.1678706258535385, + "learning_rate": 2.8221699881734e-05, + "loss": 0.029, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.13508610427379608, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0261, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12610375881195068, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0237, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.12272591143846512, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0231, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.1211056336760521, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0252, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.13204888999462128, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0276, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.11982017755508423, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0245, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.17140068113803864, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0236, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.12660211324691772, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0264, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1636001467704773, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0266, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.16939589381217957, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0276, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.15193505585193634, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0291, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16562964022159576, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0283, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.13603557646274567, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.024, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.14359496533870697, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0257, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.14077764749526978, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0247, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.14604462683200836, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0277, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.1240646094083786, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0262, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13144497573375702, + "learning_rate": 2.818639656220465e-05, + "loss": 0.0255, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.1384861171245575, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0227, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.16475403308868408, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0269, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.14310070872306824, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0258, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1498526632785797, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0263, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.13084106147289276, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0268, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.12004188448190689, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0252, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.14093606173992157, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0262, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.11016195267438889, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.0249, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.11101678013801575, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0246, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.1335095763206482, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0274, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1479913741350174, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0237, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.11868918687105179, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0231, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12799321115016937, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.027, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.17011503875255585, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0271, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.15915420651435852, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0274, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.11501206457614899, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0254, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.1457805633544922, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0274, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.1501278430223465, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0264, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.13190113008022308, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0272, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.14392758905887604, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0235, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.1375834047794342, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0241, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.14287824928760529, + "learning_rate": 2.814280752477756e-05, + "loss": 0.0269, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12234462797641754, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0259, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.1336476355791092, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0242, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.11315035820007324, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0245, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.14132073521614075, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0293, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.12053677439689636, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.0266, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.10896297544240952, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0268, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.11100424081087112, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.026, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.11256317049264908, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0237, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.14322060346603394, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0255, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.1655588448047638, + "learning_rate": 2.812283453196085e-05, + "loss": 0.026, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.14714404940605164, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0261, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.12317050993442535, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0247, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12872155010700226, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0258, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.1666097790002823, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0261, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.13425199687480927, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.16131597757339478, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0295, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.13762447237968445, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0242, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.12883082032203674, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0274, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.15029405057430267, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0269, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.13236507773399353, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0303, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.12054237723350525, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0218, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.1471109539270401, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0242, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.11926044523715973, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0275, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.13649292290210724, + "learning_rate": 2.809470494867301e-05, + "loss": 0.0231, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.15434928238391876, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0283, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.11536315083503723, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0271, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.12312797456979752, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0249, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.12750110030174255, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0235, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.1568569540977478, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0277, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.11667264997959137, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0258, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.14598897099494934, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0266, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.21723008155822754, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0258, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.14233075082302094, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0244, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.14836373925209045, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0265, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.146131694316864, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0258, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.16575932502746582, + "learning_rate": 2.807043872632495e-05, + "loss": 0.025, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.15142853558063507, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0245, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.11762700974941254, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0253, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.09714391082525253, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0233, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.1586424708366394, + "learning_rate": 2.806231820878714e-05, + "loss": 0.023, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.14317330718040466, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.0271, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.10633813589811325, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.0244, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.19832657277584076, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0249, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.14428019523620605, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0283, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.18295790255069733, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.0257, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.16204482316970825, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0285, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.14742101728916168, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0239, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1435142159461975, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0253, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.2000540792942047, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.16311310231685638, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0253, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.1238052025437355, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0249, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.16693037748336792, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0269, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.12382039427757263, + "learning_rate": 2.803581696238626e-05, + "loss": 0.025, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.13866499066352844, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0259, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.12111294269561768, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0258, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.13717366755008698, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0266, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.1119573786854744, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0264, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.1390918791294098, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.12881262600421906, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0251, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.1354016363620758, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0256, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.1285468488931656, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0315, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.17634150385856628, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0254, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.15077067911624908, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.0268, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16464342176914215, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0265, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.3138365149497986, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0255, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.15954458713531494, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0272, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15878041088581085, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0267, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.17743942141532898, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0273, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.16661205887794495, + "learning_rate": 2.800297040373295e-05, + "loss": 0.0268, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.13536596298217773, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.0241, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.13258172571659088, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0266, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.1524200588464737, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0276, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.12779775261878967, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0245, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1008654311299324, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0222, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.14413924515247345, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0297, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.14558148384094238, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0252, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.15591812133789062, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0248, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.12246204912662506, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0276, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.13106316328048706, + "learning_rate": 2.7982312890369e-05, + "loss": 0.0265, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.12663032114505768, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0234, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.1167084351181984, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15868185460567474, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0251, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13669976592063904, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0291, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.15100552141666412, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0237, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14283764362335205, + "learning_rate": 2.796987104810136e-05, + "loss": 0.024, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.13055892288684845, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0255, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.16733981668949127, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0244, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.1716628074645996, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0257, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.11838851124048233, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0288, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.14557431638240814, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0238, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.14981132745742798, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0254, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.11149631440639496, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0256, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.20500904321670532, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0255, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.12128293514251709, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0254, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.12934274971485138, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0254, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.1355874240398407, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0267, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13485173881053925, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0259, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.1481594294309616, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0251, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.15145331621170044, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0243, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.16746139526367188, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0284, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.10760390013456345, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0268, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.1460299789905548, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0264, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.15936219692230225, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0232, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.1373136043548584, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.0255, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.14508306980133057, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0243, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.153093621134758, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0262, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.15373852849006653, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0238, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.14033931493759155, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0267, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.1424216777086258, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0252, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.11820772290229797, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0253, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.13179944455623627, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0243, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.18073157966136932, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0256, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.12908786535263062, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0254, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.18455258011817932, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0263, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.14844198524951935, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0248, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.14291873574256897, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0268, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.105900838971138, + "learning_rate": 2.79029165248073e-05, + "loss": 0.027, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.14083242416381836, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0244, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.14579042792320251, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0239, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.13705924153327942, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0254, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.17981873452663422, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0225, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.13812783360481262, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0278, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12511365115642548, + "learning_rate": 2.789025068651819e-05, + "loss": 0.027, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.13009224832057953, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0245, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.145152747631073, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0281, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.1607820838689804, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0254, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.09400778263807297, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0258, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.1608203500509262, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0263, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.14968201518058777, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0247, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.1333402544260025, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0272, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.17871612310409546, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0256, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.10688412934541702, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0264, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.154439315199852, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0259, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.1270124316215515, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0279, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.15214042365550995, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0259, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.14094167947769165, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0261, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.11419129371643066, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.025, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.1447947919368744, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0258, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.12327348440885544, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0246, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.11033894866704941, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0259, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.128106951713562, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0247, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.13206088542938232, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0254, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.13664662837982178, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0274, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1349778026342392, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0262, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.10689893364906311, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0262, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.16241611540317535, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0269, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14955222606658936, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0251, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.13519948720932007, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0231, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.15076202154159546, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0272, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.15367662906646729, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0255, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.15819156169891357, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0242, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.12844862043857574, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0259, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.13392682373523712, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0245, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.15134917199611664, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.0287, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.14139734208583832, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0254, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.1562367081642151, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0246, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.1298898160457611, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0252, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13048380613327026, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0252, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.12535928189754486, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0271, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.19301816821098328, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0245, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.12262707948684692, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0262, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1394580751657486, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0272, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.21376213431358337, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0243, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.15442071855068207, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0285, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.13063938915729523, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0277, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.15023048222064972, + "learning_rate": 2.779844963973707e-05, + "loss": 0.025, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.17523249983787537, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0239, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.1312434822320938, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0234, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13352124392986298, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0249, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.12799884378910065, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0257, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.12704412639141083, + "learning_rate": 2.778765821684581e-05, + "loss": 0.025, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.11119242757558823, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0248, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.1564178764820099, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0282, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.13500694930553436, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0279, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.1542324423789978, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0258, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13888850808143616, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0281, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.14011257886886597, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.14887000620365143, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0254, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.15670792758464813, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0289, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.1288663148880005, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0242, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.13191169500350952, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.027, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.1320677250623703, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0239, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.13418279588222504, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0234, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.11019188910722733, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0232, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.13963419198989868, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0235, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.1163327619433403, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.028, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.13508521020412445, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0275, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.11098913848400116, + "learning_rate": 2.775078589916535e-05, + "loss": 0.025, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.16348809003829956, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0259, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.12986604869365692, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0246, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1302310824394226, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0259, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.11797303706407547, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.0252, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.1427793651819229, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0233, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.13411074876785278, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.0257, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.1528121531009674, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.11762527376413345, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0238, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.13639862835407257, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0257, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.1356336623430252, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0269, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.16879825294017792, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.13352762162685394, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0236, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.20271515846252441, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0296, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.11791368573904037, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0241, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14179861545562744, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.0256, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12654612958431244, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0243, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12589605152606964, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0222, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.12714968621730804, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0278, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.13410630822181702, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0248, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.12528881430625916, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0258, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.11806205660104752, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0308, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.14317870140075684, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0244, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14099767804145813, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0228, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.15186408162117004, + "learning_rate": 2.769825463131276e-05, + "loss": 0.0252, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.18532416224479675, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0262, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.1445159614086151, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0265, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14576950669288635, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0253, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1777651309967041, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0229, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.16440069675445557, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0234, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.12858232855796814, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0259, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.14329349994659424, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0254, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.1265982836484909, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0235, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13783425092697144, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0228, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.1327633559703827, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0248, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14867408573627472, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0257, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.1556752771139145, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0239, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.16407394409179688, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0261, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.14142248034477234, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0265, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.11142456531524658, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0221, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.127807155251503, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0265, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.14716863632202148, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0241, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.1455392986536026, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.024, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1665087193250656, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0253, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.16937066614627838, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0292, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.13188236951828003, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0254, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.1010027602314949, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.0236, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.15787169337272644, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.0279, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.169437438249588, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0246, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.22705122828483582, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0227, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.17403557896614075, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.027, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.1287415474653244, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0253, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.13813741505146027, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0238, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.11655540764331818, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.13267700374126434, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.025, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.10447681695222855, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0239, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.11865647882223129, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.025, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.13245069980621338, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0256, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.11377252638339996, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0266, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.11403302848339081, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0252, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.1518944650888443, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0286, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.13505733013153076, + "learning_rate": 2.761618148370765e-05, + "loss": 0.0253, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.12230793386697769, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0247, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.1600245237350464, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0267, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.13100600242614746, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.11071992665529251, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0254, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.16208025813102722, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.0287, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.1652526706457138, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.027, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14042513072490692, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.1674395203590393, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0257, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.138246089220047, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0272, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13758982717990875, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0227, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.15503156185150146, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0259, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.15050673484802246, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0279, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.13952329754829407, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.024, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.14903564751148224, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0242, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.09984514862298965, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.09952808916568756, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0233, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.14514309167861938, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0255, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.12824438512325287, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0232, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.15507018566131592, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0251, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.12279994785785675, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0225, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.13601353764533997, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0249, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.20022229850292206, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.0263, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.1407146006822586, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0226, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.1350381076335907, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0248, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.13220156729221344, + "learning_rate": 2.755998359318217e-05, + "loss": 0.0251, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.11746657639741898, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0256, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.10933379083871841, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0235, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.1263892501592636, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0244, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.14062181115150452, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.0255, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.1430615335702896, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0254, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.19069550931453705, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0254, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.16568611562252045, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0252, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.15911205112934113, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.0265, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.17054520547389984, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0265, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.15547682344913483, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0237, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.12778308987617493, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0242, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.1590787172317505, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0232, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13945519924163818, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.023, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.14507925510406494, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0247, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14541085064411163, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0245, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.17666012048721313, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0222, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.16692468523979187, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0218, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.1295796036720276, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0242, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.12527599930763245, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0273, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.13334719836711884, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0253, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.1175302267074585, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0251, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.14461147785186768, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0261, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.15658612549304962, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0236, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.15802229940891266, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0236, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.14738211035728455, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0252, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.1639637053012848, + "learning_rate": 2.750090542643057e-05, + "loss": 0.027, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1573086827993393, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0264, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.13114139437675476, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0238, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.1112227514386177, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.027, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.12209689617156982, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.13741743564605713, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0239, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14444704353809357, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0263, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.13822096586227417, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0246, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.14305464923381805, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0237, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13539855182170868, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0264, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.13791607320308685, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0234, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.205885648727417, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0234, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.1034194752573967, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0214, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.15530072152614594, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0258, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.1535874903202057, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0226, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.19778263568878174, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0238, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.175798699259758, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0255, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.1803887039422989, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0238, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.1906212419271469, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0253, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.13742655515670776, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.026, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.12772122025489807, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0261, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13805702328681946, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0259, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.14081542193889618, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0241, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.10023088753223419, + "learning_rate": 2.744810891043728e-05, + "loss": 0.0246, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.10111884027719498, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0215, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.11498598754405975, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0217, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10912947356700897, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.0257, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.13553178310394287, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0266, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.18210133910179138, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0234, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.11901862919330597, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0248, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.14622537791728973, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0234, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.1641179323196411, + "learning_rate": 2.742962757296782e-05, + "loss": 0.026, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.14369463920593262, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.0236, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.14231407642364502, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0213, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.10929716378450394, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0216, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.13811182975769043, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0241, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.13219693303108215, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0219, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.13322590291500092, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0259, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.14716853201389313, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0255, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.11750922352075577, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0272, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.1430707424879074, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1722290813922882, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.025, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.15502606332302094, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.0231, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.09124620258808136, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0246, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.16296057403087616, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0277, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.10881830751895905, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0222, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.11321897804737091, + "learning_rate": 2.739481230505444e-05, + "loss": 0.026, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.1362120658159256, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.03, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.14329490065574646, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0252, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.4512016475200653, + "learning_rate": 2.738782381719405e-05, + "loss": 0.0238, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12559202313423157, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0263, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.12616541981697083, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0248, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.1516106277704239, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0234, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.12785211205482483, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0264, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16447393596172333, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.026, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.1648014783859253, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0264, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.14414788782596588, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0249, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.10982059687376022, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0254, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.12677313387393951, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0263, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.12355965375900269, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0223, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.15190085768699646, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0246, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.13910633325576782, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0231, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.11962535977363586, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0259, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.1566285789012909, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0256, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.12282316386699677, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0253, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.16674686968326569, + "learning_rate": 2.735040900116229e-05, + "loss": 0.025, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.14292193949222565, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0277, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.1504344791173935, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0268, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.16317704319953918, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0229, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.20625433325767517, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0242, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.1895458698272705, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.024, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.1987384408712387, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.0258, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.18668580055236816, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.1807073950767517, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0246, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.11927719414234161, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.028, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.11955022811889648, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0259, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.15261328220367432, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0268, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.15620580315589905, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0254, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.12000948935747147, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0234, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.1188279539346695, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.0248, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.13279424607753754, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0241, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.13379216194152832, + "learning_rate": 2.731275407491117e-05, + "loss": 0.026, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.12054166197776794, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0215, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.15941590070724487, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0279, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.18678002059459686, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0263, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.21807490289211273, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.0239, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.12243734300136566, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.028, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.14628900587558746, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0248, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.17471152544021606, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0257, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.11189062893390656, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0242, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.16700562834739685, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0252, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.1213584691286087, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0251, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.14496786892414093, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0258, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.17899028956890106, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0248, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13594108819961548, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0246, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.1365566849708557, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.021, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.10815171152353287, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0227, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.11390877515077591, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0209, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.1541927456855774, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0213, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.1342182159423828, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0251, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.1300208568572998, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0249, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14917100965976715, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0274, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.14349427819252014, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.0269, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10514543205499649, + "learning_rate": 2.72605878418951e-05, + "loss": 0.024, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.12946246564388275, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0234, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.1120578870177269, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.0246, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.14252731204032898, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.0287, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.0984109416604042, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0234, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.13058412075042725, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0235, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.1468561738729477, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0263, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.1297929584980011, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0224, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.15071524679660797, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0254, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.15211531519889832, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0239, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.13758648931980133, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0254, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.13972750306129456, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0239, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1762881577014923, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0235, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.13277511298656464, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.025, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.11352386325597763, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0225, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.1340995877981186, + "learning_rate": 2.722476145603374e-05, + "loss": 0.0232, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.16372360289096832, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0224, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.130090594291687, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0231, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.10481512546539307, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0229, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.1453929841518402, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0255, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.133439838886261, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0248, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.1380738615989685, + "learning_rate": 2.721037237880308e-05, + "loss": 0.023, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.1386265754699707, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0264, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.14298051595687866, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.0247, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.11756420135498047, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0222, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1197751984000206, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0216, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.2194000482559204, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0243, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.14157775044441223, + "learning_rate": 2.719594991888059e-05, + "loss": 0.023, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.14389759302139282, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0228, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.12442908436059952, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0267, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.1188371554017067, + "learning_rate": 2.718872618272836e-05, + "loss": 0.027, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11137005686759949, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0237, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.10907725989818573, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0251, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.150213360786438, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0229, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.2098330855369568, + "learning_rate": 2.717908157623383e-05, + "loss": 0.0245, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.12558938562870026, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0257, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.13659170269966125, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0247, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.12037219852209091, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0228, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.13085514307022095, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0252, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.13916544616222382, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0251, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.13098886609077454, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.021, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.12777045369148254, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0258, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.10204357653856277, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0249, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.11189272999763489, + "learning_rate": 2.715732712163776e-05, + "loss": 0.0232, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.1412421613931656, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0255, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.09470678865909576, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0217, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.09736761450767517, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0233, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12988215684890747, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.1080017015337944, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0263, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.12020932137966156, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0238, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.1058862954378128, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0247, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11686211824417114, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.024, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.11078432947397232, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0226, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.13394472002983093, + "learning_rate": 2.713306780076029e-05, + "loss": 0.0237, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.11299733072519302, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.025, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.1196199283003807, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0251, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.16477037966251373, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0241, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.1175335943698883, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.11024156212806702, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0237, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.12922869622707367, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0278, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.1429518610239029, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0254, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.14210145175457, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0225, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.12270870059728622, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0257, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.12436466664075851, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0258, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.10704488307237625, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0224, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.15890415012836456, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0252, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.12372975796461105, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0225, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.11652565747499466, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.028, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.1483478546142578, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0255, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.13188686966896057, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0238, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.18725869059562683, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0244, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.15949182212352753, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0225, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.14752954244613647, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0229, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.14210951328277588, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0233, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.1388787031173706, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.0253, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.11010867357254028, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0227, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.13336122035980225, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0227, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.10118173807859421, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.0265, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.1434611827135086, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0263, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.14133326709270477, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.026, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.15359432995319366, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0236, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.12927602231502533, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.022, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.10330940783023834, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0236, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.1264553815126419, + "learning_rate": 2.705973772304842e-05, + "loss": 0.025, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.15007789433002472, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0267, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.11090201884508133, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0257, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.12769529223442078, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0252, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.0977141410112381, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0244, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.14598220586776733, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0254, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.11231301724910736, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0235, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.09610529243946075, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0262, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.1549275815486908, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0234, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.11284032464027405, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.0217, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.14465086162090302, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.0235, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.13792403042316437, + "learning_rate": 2.703264323632037e-05, + "loss": 0.024, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.1438492238521576, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0239, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.12613703310489655, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0244, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.14072397351264954, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0225, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.1420842856168747, + "learning_rate": 2.702276326859174e-05, + "loss": 0.026, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.17447218298912048, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0243, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.1708984673023224, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0259, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.1555069535970688, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.0252, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.12473946064710617, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0227, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.15750949084758759, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0248, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.14009647071361542, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0267, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.1416911780834198, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0229, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.11944608390331268, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0235, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.126204252243042, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0268, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.1186814159154892, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0242, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.12699876725673676, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0239, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.10452742129564285, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0221, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.1422647386789322, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0247, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.11601094901561737, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0232, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12339281290769577, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0223, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1883407086133957, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0243, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.1730756312608719, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0294, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.10429747402667999, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0194, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.1358909010887146, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0217, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.14417894184589386, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0265, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.12299799919128418, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0236, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.1361788511276245, + "learning_rate": 2.696816263672279e-05, + "loss": 0.024, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.14492252469062805, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0241, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.1209225058555603, + "learning_rate": 2.696317709902802e-05, + "loss": 0.0239, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.11048539727926254, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.12423776835203171, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0243, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.1254356950521469, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.12912537157535553, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0254, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.10864657163619995, + "learning_rate": 2.695069735893089e-05, + "loss": 0.022, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.15519478917121887, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0253, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.15276901423931122, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0231, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.1401076465845108, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0271, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.14237190783023834, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.024, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1186559721827507, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0238, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.1254940629005432, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0233, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.1337885558605194, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0255, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12958411872386932, + "learning_rate": 2.69306825906724e-05, + "loss": 0.025, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.11610005795955658, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0238, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.13515442609786987, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0239, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1492868959903717, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0224, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.1293300986289978, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.021, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12354710698127747, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.13226614892482758, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.0247, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.1368253529071808, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0255, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.13044875860214233, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0228, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14107753336429596, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0221, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.13521252572536469, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0236, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.13718804717063904, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0217, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.15192179381847382, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.0247, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.13718710839748383, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0244, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11781664192676544, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0239, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.14431379735469818, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0248, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.15059512853622437, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0231, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.12076479196548462, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0242, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.19006308913230896, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0238, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.140592560172081, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0233, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.1528412401676178, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0229, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.16418038308620453, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0242, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.12185576558113098, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0254, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11594245582818985, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0248, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1216709092259407, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0234, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.12024594098329544, + "learning_rate": 2.686776313743295e-05, + "loss": 0.024, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.1321716010570526, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.0255, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.13785192370414734, + "learning_rate": 2.686270520700719e-05, + "loss": 0.025, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.12321551889181137, + "learning_rate": 2.6860174890357e-05, + "loss": 0.0233, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14606279134750366, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0249, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.17512187361717224, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0272, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.14974310994148254, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.0243, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.11758870631456375, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0241, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.11157143115997314, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0232, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.10899091511964798, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0231, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.1136092096567154, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0224, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.10306857526302338, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0211, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13241226971149445, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0222, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.11886361986398697, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0244, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.11962216347455978, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0237, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.14516247808933258, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.14592237770557404, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0253, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1278424859046936, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0233, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.12451537698507309, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0218, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.11746378242969513, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0266, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.14618739485740662, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0226, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.1512313336133957, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0233, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.14035750925540924, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0264, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.13232699036598206, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.023, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.14616237580776215, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.0233, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.12882372736930847, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0265, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.13778620958328247, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.024, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.15017054975032806, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0239, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.14781621098518372, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0215, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1670294851064682, + "learning_rate": 2.679407107451536e-05, + "loss": 0.022, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.1112075075507164, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0232, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.12219138443470001, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0202, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12922540307044983, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.11967732757329941, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0226, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.1377563774585724, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0224, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.14771930873394012, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0234, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.14800170063972473, + "learning_rate": 2.677617030053916e-05, + "loss": 0.0251, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.1233784481883049, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0232, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.16905057430267334, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0214, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.13751909136772156, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0248, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.13289029896259308, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0216, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.10788674652576447, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0216, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.13367260992527008, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1270209699869156, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0239, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.140412375330925, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0218, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.17788565158843994, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0253, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.13695330917835236, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.023, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.1250872015953064, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0232, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.11347977817058563, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0243, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.15546919405460358, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0233, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.13752058148384094, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.10431966185569763, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0215, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.17878347635269165, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0262, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.13218404352664948, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0241, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.12244930118322372, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.0247, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.11582386493682861, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0219, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.1160157322883606, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0203, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.10731764137744904, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0234, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.11152671277523041, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0218, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.12760375440120697, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0218, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.13706916570663452, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0219, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.14939993619918823, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0208, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.1169784739613533, + "learning_rate": 2.67092985767871e-05, + "loss": 0.0217, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.1569128930568695, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12233320623636246, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0251, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13152818381786346, + "learning_rate": 2.670154386126402e-05, + "loss": 0.0231, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.1529632955789566, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0246, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.11001774668693542, + "learning_rate": 2.669636960731296e-05, + "loss": 0.0223, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.1473267376422882, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.025, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.12853027880191803, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0246, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.15698745846748352, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0209, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12175270169973373, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0229, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.15634675323963165, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.0224, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.16060872375965118, + "learning_rate": 2.668082553346281e-05, + "loss": 0.0254, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.13773131370544434, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0224, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.14258436858654022, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0216, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.13032476603984833, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0215, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.17300327122211456, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0221, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.13187018036842346, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0226, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.14065369963645935, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0232, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.11484711617231369, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0263, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1380661576986313, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0209, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.17800134420394897, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0244, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.14836716651916504, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0277, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14563506841659546, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.023, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.11463280767202377, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.026, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.13092686235904694, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0222, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.12265385687351227, + "learning_rate": 2.664443191002912e-05, + "loss": 0.025, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.14121955633163452, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0242, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.12938837707042694, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0222, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.1922816038131714, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.0228, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.12596914172172546, + "learning_rate": 2.663400187275054e-05, + "loss": 0.022, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.14095240831375122, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0244, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.1080891340970993, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0228, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.13916714489459991, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.0239, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.1489475965499878, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0251, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.11781706660985947, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0207, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.11544635146856308, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0232, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.14827974140644073, + "learning_rate": 2.661571529941175e-05, + "loss": 0.0236, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.12821446359157562, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.0186, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.1303764134645462, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0234, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.1049366369843483, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0232, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.15925322473049164, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0275, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.13151033222675323, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0213, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.09425143152475357, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0215, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.15108847618103027, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0228, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.1374080330133438, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0231, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.16061817109584808, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0253, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.1707434207201004, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0259, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.12051897495985031, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0217, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.09191743284463882, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0224, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.12547235190868378, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0245, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14491257071495056, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0228, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.11896242201328278, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0239, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.14253611862659454, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0233, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.12803201377391815, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.0226, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.12893305718898773, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0231, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.16127154231071472, + "learning_rate": 2.656586262343925e-05, + "loss": 0.0234, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.13307608664035797, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0258, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.11211924254894257, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.0219, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.1672789305448532, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0248, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.12878142297267914, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0231, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.11429315060377121, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0233, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.13469499349594116, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0254, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.12753085792064667, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0258, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.11875704675912857, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0234, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.13718342781066895, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0223, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.10762278735637665, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.021, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.11651597917079926, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0222, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.12953802943229675, + "learning_rate": 2.653421326308085e-05, + "loss": 0.0235, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.14820824563503265, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.023, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.16258545219898224, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0266, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.14097236096858978, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0224, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.1039188802242279, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0216, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15014079213142395, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0246, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12850438058376312, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.023, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.12245986610651016, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0209, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.15426407754421234, + "learning_rate": 2.651304359235998e-05, + "loss": 0.025, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.13411583006381989, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0229, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.13705122470855713, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0254, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.13120216131210327, + "learning_rate": 2.650509053370553e-05, + "loss": 0.022, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14599482715129852, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0249, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.11628340184688568, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0215, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.13643144071102142, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0228, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.12846967577934265, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0214, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.13181492686271667, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0245, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1277213841676712, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0244, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12288910150527954, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0226, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.14789454638957977, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0215, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.11457780748605728, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0248, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.12511487305164337, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0206, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.12255935370922089, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0245, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.13214053213596344, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0238, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.1601077914237976, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.027, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.1244364008307457, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0212, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.11869437992572784, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0239, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.14327670633792877, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0227, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.12729141116142273, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0248, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.10953658074140549, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0205, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.20326738059520721, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0225, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.15614672005176544, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0237, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.11863692849874496, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0208, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.10683806240558624, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0224, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.1529647260904312, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.024, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.18428578972816467, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0232, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.19935141503810883, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.0223, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.14675194025039673, + "learning_rate": 2.643583521046136e-05, + "loss": 0.0216, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.21206484735012054, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0243, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.14521193504333496, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0214, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.11565426737070084, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0236, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.1087750792503357, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0215, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.11942578107118607, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0229, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.138235405087471, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0219, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.1079302579164505, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0222, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.15928085148334503, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0225, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.10920153558254242, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.024, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.12247785925865173, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0205, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.10454121232032776, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0238, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.1726839691400528, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0211, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1617811918258667, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0225, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.15452679991722107, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0214, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1573309749364853, + "learning_rate": 2.639561293166059e-05, + "loss": 0.022, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.13596323132514954, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0234, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.09188699722290039, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0219, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.12158972024917603, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0202, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.16615170240402222, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.0221, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.1163380891084671, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0255, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.17704762518405914, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0243, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.1591942012310028, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.021, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.14697043597698212, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0241, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1234351247549057, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0261, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.1379440575838089, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0266, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1025586649775505, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.022, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.11894206702709198, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0226, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1431341916322708, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0241, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.10291697829961777, + "learning_rate": 2.635789644394928e-05, + "loss": 0.022, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.1622016727924347, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0229, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.11867184937000275, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0262, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.18177814781665802, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0237, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.15159954130649567, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0218, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.11481288820505142, + "learning_rate": 2.634438526911474e-05, + "loss": 0.024, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.10989677160978317, + "learning_rate": 2.634168044873305e-05, + "loss": 0.024, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.12370797991752625, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0235, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1390318125486374, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0231, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.14030151069164276, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.0207, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11659140884876251, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.021, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.16727864742279053, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0242, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.11524021625518799, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0265, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.10776443779468536, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0224, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.16366012394428253, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0264, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.1102370023727417, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0245, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11543556302785873, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0216, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.11101235449314117, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0208, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.1257167011499405, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.024, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.1063276082277298, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0241, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.14303457736968994, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0242, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.12862786650657654, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0244, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15487778186798096, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.14244592189788818, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0211, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1066693440079689, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0199, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11721482872962952, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0214, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.13803230226039886, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.0227, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.11225558817386627, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.028, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.11660541594028473, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0239, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12551093101501465, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0221, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.1305139809846878, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.1281510442495346, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0235, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.14735114574432373, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0213, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.10699455440044403, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0231, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.16579580307006836, + "learning_rate": 2.626559653214893e-05, + "loss": 0.0213, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.09389153122901917, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0218, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.17565062642097473, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0245, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.1368194818496704, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0214, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.10599412024021149, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0221, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.12645669281482697, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0233, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.12325828522443771, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.0228, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.11353788524866104, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.10889376699924469, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0227, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.12935040891170502, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0214, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.1422247439622879, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0219, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.16723784804344177, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0239, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.09775881469249725, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.0229, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.13455381989479065, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0253, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.13700921833515167, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0239, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12252738326787949, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0231, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.15983854234218597, + "learning_rate": 2.622181856488204e-05, + "loss": 0.0223, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.18716242909431458, + "learning_rate": 2.621907518918971e-05, + "loss": 0.025, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.16380536556243896, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0229, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.1524866819381714, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0221, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.12281060218811035, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0213, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.09793791174888611, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0249, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.12271476536989212, + "learning_rate": 2.620534553467634e-05, + "loss": 0.0221, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.12616175413131714, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0236, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.10274055600166321, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0205, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.16873067617416382, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.027, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.10157257318496704, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0236, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.11863094568252563, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0225, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.09981630742549896, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0233, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11045375466346741, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0226, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.13111108541488647, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0232, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.12417030334472656, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0226, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.12029916048049927, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0217, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.10766672343015671, + "learning_rate": 2.617506544598029e-05, + "loss": 0.025, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.12039721757173538, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0228, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.11044551432132721, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0223, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.12463348358869553, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.0232, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.14382848143577576, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0238, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.13677646219730377, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0239, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.11536712944507599, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0224, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.12461430579423904, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0253, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.16901540756225586, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0242, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.12255802005529404, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0212, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.14185196161270142, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.0218, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.1292722374200821, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0249, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12001507729291916, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0222, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.12286248803138733, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0211, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.12487491965293884, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0231, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13417202234268188, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0257, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.11690660566091537, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0219, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15474916994571686, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0223, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.1720103919506073, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0239, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.16786889731884003, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0216, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.2279045283794403, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.0254, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13872499763965607, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0238, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.1881643831729889, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0211, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.1372661143541336, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0237, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.10514581203460693, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0217, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.11103228479623795, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0248, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1349911093711853, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0237, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.12774604558944702, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0245, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.15125125646591187, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0233, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.1190832257270813, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0248, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.1088283434510231, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0187, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.12167634069919586, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.023, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.15135347843170166, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0226, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.12857669591903687, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.0195, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.13588359951972961, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0234, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.14902308583259583, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0235, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.15095068514347076, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0209, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.1116727739572525, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0255, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11652971059083939, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0211, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.11875079572200775, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0223, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.11100098490715027, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.022, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.13648055493831635, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0234, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.12870723009109497, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0205, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.10456091910600662, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0232, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.13770641386508942, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0212, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.1526155173778534, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0244, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12664297223091125, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0241, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.12000498175621033, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0239, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.11930232495069504, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.0207, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.11384909600019455, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0217, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.1746685951948166, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.023, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.15868088603019714, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0233, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.15131928026676178, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0194, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.10279352962970734, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0215, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.13810355961322784, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0202, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.13024795055389404, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0217, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1374605894088745, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0215, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.14356477558612823, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0206, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.11046172678470612, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0236, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.10146678239107132, + "learning_rate": 2.601090850395125e-05, + "loss": 0.021, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13247323036193848, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0217, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.13672757148742676, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0201, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.08651759475469589, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.0196, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.14821568131446838, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0223, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11985234171152115, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0208, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11231720447540283, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0219, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.12338588386774063, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0202, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.15079368650913239, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0229, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.1348753422498703, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0209, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1333470344543457, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0207, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.10075798630714417, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0232, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1217522993683815, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0247, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.12041056901216507, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.022, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.10997799783945084, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0226, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.119750015437603, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0239, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.15246307849884033, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0239, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.1431925892829895, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0227, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.1181643158197403, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0227, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.13645906746387482, + "learning_rate": 2.595742267991401e-05, + "loss": 0.0213, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.12661530077457428, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0233, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.1344967484474182, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.0221, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.13433857262134552, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0238, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.10517163574695587, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0237, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.10689525306224823, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.022, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.13307900726795197, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0209, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.14696283638477325, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0261, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.12473653256893158, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0213, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.106630340218544, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0221, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.12092334777116776, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0232, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.11033511906862259, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0193, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11916322261095047, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0212, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.1204444020986557, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0228, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.1373489499092102, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0225, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.13124211132526398, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0208, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.18858365714550018, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0258, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.1558879315853119, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0228, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.1481737494468689, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0242, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.10424906760454178, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.021, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.15494640171527863, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0238, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.1801268309354782, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0238, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.14176428318023682, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0228, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.15850430727005005, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0235, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.14727924764156342, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0232, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.14907293021678925, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0242, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.11011895537376404, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0231, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.13541805744171143, + "learning_rate": 2.588090010443142e-05, + "loss": 0.0224, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.13078846037387848, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0228, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.14508767426013947, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0207, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.12445177882909775, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0227, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.14860141277313232, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0244, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.13311992585659027, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0207, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.1157015860080719, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0209, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.13310302793979645, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.0234, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.14026974141597748, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0246, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.14788632094860077, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0213, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.1551811844110489, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0241, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.10706475377082825, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.021, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.12536053359508514, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0228, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.1260751187801361, + "learning_rate": 2.584384079351628e-05, + "loss": 0.022, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.1082935482263565, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0207, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.13922622799873352, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0251, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1240227147936821, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0223, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.14512665569782257, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0218, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12859801948070526, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.0208, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.13418391346931458, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.021, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.10546918958425522, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.022, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.10668392479419708, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10407276451587677, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.02, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.13980497419834137, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0235, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.16233955323696136, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0224, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.16743852198123932, + "learning_rate": 2.580950866038058e-05, + "loss": 0.022, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.09152300655841827, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0206, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.14319349825382233, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0205, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.1045396700501442, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0219, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.10865644365549088, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0201, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.12166012078523636, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0217, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.12505589425563812, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0224, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.1353408247232437, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0242, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.1379590630531311, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0221, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12337373197078705, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0235, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.12092723697423935, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0233, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.09944337606430054, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0215, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.15161453187465668, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0193, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.0963459238409996, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0225, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.15534570813179016, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0247, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.142191082239151, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0228, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.1454230695962906, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0226, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.12661893665790558, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0211, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.11373333632946014, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0213, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12031935155391693, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.0208, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.13401862978935242, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0224, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.11285468935966492, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0231, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.12453637272119522, + "learning_rate": 2.574625966768285e-05, + "loss": 0.0211, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.13041555881500244, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1391122043132782, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0224, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.1345762014389038, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0244, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.12814867496490479, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0219, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.11058632284402847, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0191, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.15668845176696777, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.021, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.1509203314781189, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.0254, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.13209059834480286, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0201, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.1357070952653885, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0226, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.0896834284067154, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0207, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.1570383608341217, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0234, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.16916486620903015, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0254, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.1385425180196762, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.0218, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.11236828565597534, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0214, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11224918812513351, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0233, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.1305687576532364, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0219, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.15051157772541046, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0226, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.13418416678905487, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0244, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.14324583113193512, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0235, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.10810746997594833, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0197, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.14513953030109406, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0215, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1187531128525734, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0193, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11865831911563873, + "learning_rate": 2.567971342427169e-05, + "loss": 0.021, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.12040061503648758, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0229, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.10597001016139984, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0209, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.1219123974442482, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0242, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.11732123047113419, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.0203, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.15642790496349335, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0226, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.1893257051706314, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0247, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11829031258821487, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0239, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.14971555769443512, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0253, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.12424097210168839, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0251, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1071329414844513, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0271, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.11146610230207443, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0214, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.10925749689340591, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.0212, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.14797845482826233, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0221, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.10348086059093475, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0218, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.14607182145118713, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0217, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.11097954958677292, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0259, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.11300431936979294, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0197, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.11342521011829376, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0202, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.12889035046100616, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0217, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.10449834913015366, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0216, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1293669193983078, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0227, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.10667148977518082, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.021, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.0839434340596199, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0236, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.12906265258789062, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0223, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.1100582480430603, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0243, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.10644520074129105, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.0215, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.11693669110536575, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.022, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.13665053248405457, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0227, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12752027809619904, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0241, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.11534811556339264, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0234, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.1467696875333786, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0245, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10596702992916107, + "learning_rate": 2.558641420481684e-05, + "loss": 0.02, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.14685864746570587, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.022, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.13554386794567108, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.026, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.12572740018367767, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0232, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.1583555042743683, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0209, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.12913484871387482, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0226, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.11418721824884415, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0225, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.12722279131412506, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.0211, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.1382816582918167, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.0243, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.11612103134393692, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0212, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.1537163406610489, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.021, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.10462892800569534, + "learning_rate": 2.555415209239832e-05, + "loss": 0.0214, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.12759479880332947, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0232, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.1290607452392578, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0222, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12458004057407379, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0213, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.1113182008266449, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0223, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.15024979412555695, + "learning_rate": 2.553945541728904e-05, + "loss": 0.022, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.12657828629016876, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.025, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.13719463348388672, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.024, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.13465221226215363, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0254, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.08891251683235168, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0215, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.08946283161640167, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0224, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.13241387903690338, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0229, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.13068708777427673, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0215, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.12115749716758728, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.025, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.12863430380821228, + "learning_rate": 2.551295099554365e-05, + "loss": 0.0202, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.1332341581583023, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0204, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.102240651845932, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0199, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.16963273286819458, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0226, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12840960919857025, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0225, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.09958021342754364, + "learning_rate": 2.549819836273345e-05, + "loss": 0.02, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.10737278312444687, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0226, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.11473958194255829, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0219, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.12188851088285446, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0216, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.1229034960269928, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0227, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.1316317766904831, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1230575442314148, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0218, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13473588228225708, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0221, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1293632537126541, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0223, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.11663107573986053, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0198, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.147003173828125, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0216, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.10070289671421051, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0204, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12855689227581024, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0222, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.10310211777687073, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0227, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.25780248641967773, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0204, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.11621664464473724, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0196, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.15240156650543213, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0232, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.1359490305185318, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0191, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.13072653114795685, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.0242, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.1296282559633255, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0214, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.12183726578950882, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0215, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.10602327436208725, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0222, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.12424091249704361, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0215, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.14287705719470978, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0217, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.10017845034599304, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0214, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.1265304833650589, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0214, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.15279516577720642, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0194, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.11116814613342285, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0197, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.14380785822868347, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0187, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.12522602081298828, + "learning_rate": 2.541224092500406e-05, + "loss": 0.023, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.12386767566204071, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.0197, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11278095096349716, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0205, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12119927257299423, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0206, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.14128951728343964, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0197, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.12576138973236084, + "learning_rate": 2.539735331078111e-05, + "loss": 0.024, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.13276934623718262, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0217, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.15223100781440735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0222, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.13918957114219666, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0211, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.10129881650209427, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0192, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.15048135817050934, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0191, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.11579657346010208, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.019, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.1882474422454834, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0241, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.12608706951141357, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.021, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.14131402969360352, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0234, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.11546603590250015, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0208, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.0989922359585762, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0219, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.13154740631580353, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0234, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.12450987845659256, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0219, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.15696841478347778, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0228, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12059786915779114, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0236, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.1058221235871315, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0234, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.11429761350154877, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0232, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.11841994524002075, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0193, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.12387857586145401, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0239, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.11453676968812943, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0224, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.15571972727775574, + "learning_rate": 2.533461009735874e-05, + "loss": 0.022, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.10241609811782837, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.022, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.11682361364364624, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0211, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.13623486459255219, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0215, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.12484022229909897, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0212, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.09105533361434937, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0223, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.1024160087108612, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0174, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.14108790457248688, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0215, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.1268249899148941, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0211, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12255556881427765, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0192, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.15904900431632996, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0259, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.1356613039970398, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.023, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.12410393357276917, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0223, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1291753053665161, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0221, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.109084852039814, + "learning_rate": 2.529258887162501e-05, + "loss": 0.0243, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.16595998406410217, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0201, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.14062269032001495, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0232, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.13198593258857727, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0214, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.10668856650590897, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0229, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.11266572028398514, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0233, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.12742944061756134, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0221, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.12420892715454102, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0215, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.1265600174665451, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0215, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.11914686858654022, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0222, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.12876294553279877, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0233, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.09800750762224197, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0219, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1273864060640335, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0238, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.13299106061458588, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0226, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.09396592527627945, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.12955208122730255, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0204, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.16461864113807678, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0211, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.11535970121622086, + "learning_rate": 2.524135719213194e-05, + "loss": 0.02, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.11860322207212448, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0212, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10968036204576492, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0207, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.12494095414876938, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.02, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.10398012399673462, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0205, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.12939494848251343, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0216, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.10996326059103012, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0199, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.13704033195972443, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0208, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.11454308778047562, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.0212, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.14186057448387146, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0226, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.12624409794807434, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0199, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.14392296969890594, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.15774424374103546, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0213, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.1352480947971344, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0192, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.10594955086708069, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0231, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.11280091851949692, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0202, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11683343350887299, + "learning_rate": 2.519293379145164e-05, + "loss": 0.022, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.11721377819776535, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0223, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.18298691511154175, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.0225, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12927964329719543, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0202, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.1413726508617401, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.021, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.1330733746290207, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0228, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12341104447841644, + "learning_rate": 2.517472387591168e-05, + "loss": 0.021, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.12199880182743073, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.0202, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.12259527295827866, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.0227, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.11682213097810745, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0246, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.12783272564411163, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0204, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.09663784503936768, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0228, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.13686567544937134, + "learning_rate": 2.515648614306516e-05, + "loss": 0.02, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.12164216488599777, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0234, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.1371183544397354, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0217, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.12031334638595581, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0203, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.1183433011174202, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.1265946477651596, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0224, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.12686678767204285, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0202, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.13161398470401764, + "learning_rate": 2.513517369661726e-05, + "loss": 0.0216, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.13780610263347626, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0205, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.14213374257087708, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0237, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.12532930076122284, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0215, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1276065707206726, + "learning_rate": 2.512297821731217e-05, + "loss": 0.02, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.08893107622861862, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12819649279117584, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0202, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.09707193821668625, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0202, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.511077043763402e-05, + "loss": 0.0208, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.10047636926174164, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0219, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.15637417137622833, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0218, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.15387912094593048, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0213, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.14876650273799896, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0222, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.13631325960159302, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0214, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.12577500939369202, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0232, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15896311402320862, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0237, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.16459719836711884, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0223, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11081109195947647, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0223, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.12327340245246887, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0219, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14812195301055908, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0228, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.11862681806087494, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0252, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.10684223473072052, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0227, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.1279568374156952, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0216, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.13341103494167328, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0213, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.11913862824440002, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0214, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.13493330776691437, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.0214, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.1272144317626953, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0209, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.11348695307970047, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0217, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.16526950895786285, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0198, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.14953579008579254, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0231, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.12980960309505463, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0226, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.16652929782867432, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0244, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1722235530614853, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.026, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.23629626631736755, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0221, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.12417787313461304, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.023, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.11768155544996262, + "learning_rate": 2.502804732645032e-05, + "loss": 0.0186, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.13731959462165833, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.023, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.18034201860427856, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.0201, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11240512877702713, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0217, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.10463474690914154, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0201, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.1273874044418335, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0194, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.12314468622207642, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0237, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.13170552253723145, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.021, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.12725704908370972, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0198, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.11761748790740967, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0198, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13019445538520813, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0237, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.09089519083499908, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.0202, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.13669925928115845, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0236, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.0891583263874054, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0189, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.12015428394079208, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.0214, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.12889859080314636, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0214, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.1092396154999733, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.022, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.1011805608868599, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0201, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.11652060598134995, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.019, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.12280789762735367, + "learning_rate": 2.496950169240971e-05, + "loss": 0.0229, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.1253744661808014, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0245, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.15854160487651825, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0257, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.11671199649572372, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0204, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.15195102989673615, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.11878954619169235, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0207, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.1035882979631424, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0195, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12105780094861984, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0191, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.13278989493846893, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0207, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.16576215624809265, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.0228, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.13104894757270813, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0197, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.14476071298122406, + "learning_rate": 2.49354818242132e-05, + "loss": 0.023, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.14116041362285614, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0204, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.133507639169693, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0213, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11147522926330566, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0192, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.1317354440689087, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0229, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.15885432064533234, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0214, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.0996033102273941, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0213, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.11059842258691788, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0181, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.1103382259607315, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0252, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.12147006392478943, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.1167205348610878, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0217, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.11838674545288086, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0223, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.1745499074459076, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0213, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.09892687201499939, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0213, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.17135944962501526, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0228, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.13930797576904297, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0211, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.10764696449041367, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0204, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11211943626403809, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0207, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1359340399503708, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0215, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.11819283664226532, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0213, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.1166793555021286, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0216, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.14411170780658722, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0206, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.1380920261144638, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0211, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.12901239097118378, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.0224, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.1171313151717186, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.0207, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.15169396996498108, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0225, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.2066943198442459, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0182, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1179543286561966, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0226, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.12273375689983368, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.0224, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.09459412097930908, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0177, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.10208278894424438, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0207, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.11947600543498993, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.0246, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10635806620121002, + "learning_rate": 2.483599695058918e-05, + "loss": 0.021, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.15165211260318756, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0242, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.10573819279670715, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0221, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.10153646767139435, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0212, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.12596352398395538, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.021, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.11259406059980392, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0214, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.09709548950195312, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0186, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1462256908416748, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0228, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.15480898320674896, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0245, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.10707002878189087, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0197, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.10963082313537598, + "learning_rate": 2.480475069778109e-05, + "loss": 0.019, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.13645771145820618, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0229, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.08208916336297989, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0194, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.11223483830690384, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0212, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.12233787775039673, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0228, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.12175288051366806, + "learning_rate": 2.478909963381241e-05, + "loss": 0.0204, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.14267505705356598, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0203, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.12649142742156982, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0217, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.11254285275936127, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0214, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.1454164832830429, + "learning_rate": 2.477656539969673e-05, + "loss": 0.0221, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.17177407443523407, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0212, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.1089746281504631, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0222, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.1525813490152359, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0236, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.09651731699705124, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0214, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.12710145115852356, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0233, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.15999987721443176, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0215, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1600969433784485, + "learning_rate": 2.475460191424065e-05, + "loss": 0.0201, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.11045356094837189, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0208, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.14192573726177216, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.0216, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.11450321227312088, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0202, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.13454747200012207, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.021, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.11518846452236176, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0194, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.10673099011182785, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0214, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.12131883949041367, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0203, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.18098120391368866, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0182, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.09981405735015869, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0214, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.09194957464933395, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0209, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.11133445054292679, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0227, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.14570456743240356, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0181, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.14878204464912415, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0213, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.1475936472415924, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0229, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.14091694355010986, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0229, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.10411489754915237, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.021, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.10015816241502762, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0211, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09339930862188339, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0206, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.13016946613788605, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0223, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.10682770609855652, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0258, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.13658446073532104, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0232, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.12742926180362701, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.13928014039993286, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0208, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.10824346542358398, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0199, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.11346063017845154, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.0225, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12237603217363358, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0226, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.12201637774705887, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0209, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.09991955012083054, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0239, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.15387621521949768, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.025, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.1588744819164276, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0192, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.1347019523382187, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0196, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.12746664881706238, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0212, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11522521078586578, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0197, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.09697534143924713, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0219, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12183485180139542, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0195, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.13119301199913025, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0195, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.10748504102230072, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0192, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.1617487072944641, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0224, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.11518653482198715, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0207, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.17646238207817078, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0233, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.13729365170001984, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.09777814894914627, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0207, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.117672860622406, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0254, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.09014946967363358, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0214, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.1273731142282486, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0223, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10932891070842743, + "learning_rate": 2.460937119575347e-05, + "loss": 0.0224, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.1672651767730713, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0205, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.0930962786078453, + "learning_rate": 2.46030216185615e-05, + "loss": 0.02, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.09708375483751297, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0192, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.13240690529346466, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0214, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.10310725122690201, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0204, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.13352994620800018, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0206, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.15277482569217682, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0222, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.14223086833953857, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0229, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12845414876937866, + "learning_rate": 2.458077513854331e-05, + "loss": 0.0221, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13195961713790894, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0232, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.1461530476808548, + "learning_rate": 2.457441244940383e-05, + "loss": 0.022, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13729079067707062, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0196, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.15615087747573853, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0213, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.12907880544662476, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0198, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.11320316791534424, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0225, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.14982816576957703, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0225, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.12824060022830963, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0196, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.11940523982048035, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0204, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946357995271683, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.0218, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.13296495378017426, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0229, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.1383914053440094, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0207, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.11321834474802017, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0202, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.1276783049106598, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0193, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1406320035457611, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0191, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.12060298770666122, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0195, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.08627335727214813, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0191, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1429424285888672, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0197, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.15133187174797058, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0212, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.11892743408679962, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0231, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.11871744692325592, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0211, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.105750173330307, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.0234, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10969540476799011, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.022, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.09607745707035065, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0173, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.11796857416629791, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0193, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.114372618496418, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.02, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.1164821982383728, + "learning_rate": 2.449463406055895e-05, + "loss": 0.023, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.13561253249645233, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.018, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.14956215023994446, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0245, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.12785622477531433, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0222, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.16085495054721832, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.021, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10792174935340881, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0209, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.10508404672145844, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0217, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.11471311002969742, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0202, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.14265185594558716, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0224, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.13569262623786926, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0196, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15635624527931213, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0205, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12042741477489471, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0179, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.09488346427679062, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0198, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.20503900945186615, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0199, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.13667535781860352, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13254515826702118, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0204, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.13879507780075073, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0233, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.11448678374290466, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0237, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1255042403936386, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0204, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11896912008523941, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0207, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.12624827027320862, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0194, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.09312308579683304, + "learning_rate": 2.442727178914008e-05, + "loss": 0.022, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12632645666599274, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0189, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.11468527466058731, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0182, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.09245705604553223, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0188, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.12113099545240402, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0191, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.13049040734767914, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0221, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.16979938745498657, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0211, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.11175577342510223, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0172, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.1191241443157196, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0206, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.16317729651927948, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.0213, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.10087717324495316, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0194, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.12195733189582825, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0253, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11880146712064743, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0222, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.14213789999485016, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0198, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.15624478459358215, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.0205, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09695806354284286, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0219, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11260177940130234, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0206, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.15394112467765808, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0219, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.11663617193698883, + "learning_rate": 2.436928131964383e-05, + "loss": 0.0214, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.12596042454242706, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0225, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.10310690104961395, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0203, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.12332364916801453, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0203, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.16257967054843903, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0234, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.13102053105831146, + "learning_rate": 2.435313188161029e-05, + "loss": 0.022, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.14736537635326385, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0199, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.14611469209194183, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0238, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.14594200253486633, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0212, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.16921328008174896, + "learning_rate": 2.434019954415699e-05, + "loss": 0.0236, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.1318638175725937, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0201, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.10869060456752777, + "learning_rate": 2.433372911848899e-05, + "loss": 0.019, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.1446760594844818, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.021, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.1298162341117859, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10307328402996063, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0197, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.11932921409606934, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0209, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.10986137390136719, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0227, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.12099824845790863, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0182, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.13432320952415466, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.0213, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.14018790423870087, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0199, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.13015322387218475, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0191, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.10281360149383545, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0215, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.09500012546777725, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0216, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.15131708979606628, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0225, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.08750247210264206, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0221, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.10143480449914932, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0214, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.12935201823711395, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0202, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.10703578591346741, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0207, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08649012446403503, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0184, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.11186458170413971, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.021, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.11181459575891495, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0211, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.11756999045610428, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0216, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.12061341106891632, + "learning_rate": 2.426561884159603e-05, + "loss": 0.019, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.13861820101737976, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0214, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.11545911431312561, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0195, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.14892302453517914, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0226, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.10354511439800262, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0197, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.15290679037570953, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0208, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.12185964733362198, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0205, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.10486926138401031, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0206, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.10959216207265854, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0198, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.11638635396957397, + "learning_rate": 2.423633359536448e-05, + "loss": 0.0227, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.13661672174930573, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0206, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.13071168959140778, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0189, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.16377446055412292, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0203, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.22221329808235168, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0185, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.11251460015773773, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0191, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.10873523354530334, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0228, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.13705629110336304, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0203, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.13980528712272644, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.10355913639068604, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0199, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.1282302737236023, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.0215, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.1221374049782753, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0192, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.18542620539665222, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0199, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1250954121351242, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0189, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12665164470672607, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0217, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.09464617818593979, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.0197, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.09657798707485199, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0194, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.11533467471599579, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0189, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.13647402822971344, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0194, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.13276426494121552, + "learning_rate": 2.41743228236291e-05, + "loss": 0.022, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.16899874806404114, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0209, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.14313359558582306, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0211, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.14507170021533966, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0218, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.13025349378585815, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0209, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.1666184365749359, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0217, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14891882240772247, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0205, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11392978578805923, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0223, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.12964752316474915, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.019, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11125730723142624, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0211, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.08378217369318008, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0203, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.14530996978282928, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0206, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.10187243670225143, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0188, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.09652446955442429, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0192, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.10552513599395752, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0212, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11044801771640778, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0215, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.12248426675796509, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0185, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.11164992302656174, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0198, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.11919678002595901, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0187, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1675145924091339, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0262, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.15451745688915253, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0226, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.0997307077050209, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0196, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.1272345781326294, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0214, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.1286081224679947, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0214, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12226564437150955, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0211, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.11795010417699814, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0216, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.12963736057281494, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0215, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.10415369272232056, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0215, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.11642655730247498, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0194, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.19006073474884033, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0216, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.16677096486091614, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.0235, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.15928123891353607, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0195, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.10270075500011444, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0197, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.12253808975219727, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0216, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14145877957344055, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0211, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1387409120798111, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0185, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.11433693021535873, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0204, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.14247572422027588, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0187, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.12203793972730637, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.018, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.15705470740795135, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0212, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.10648002475500107, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0182, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.11573450267314911, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0211, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.11960671842098236, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0245, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.14416365325450897, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0178, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.11234421283006668, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0205, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.1375260204076767, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.025, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.08686859905719757, + "learning_rate": 2.402315322351119e-05, + "loss": 0.018, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11352542787790298, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0198, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.13374295830726624, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0216, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.09358088672161102, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0202, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.13767337799072266, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0229, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.1405765861272812, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.02, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11317892372608185, + "learning_rate": 2.400332803040664e-05, + "loss": 0.02, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.12356667965650558, + "learning_rate": 2.400002143698994e-05, + "loss": 0.0225, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.12570610642433167, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0232, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.1461946964263916, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0201, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.10893318802118301, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0211, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.10969080775976181, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0235, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.17320629954338074, + "learning_rate": 2.398347822254704e-05, + "loss": 0.0221, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14296430349349976, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0206, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12349261343479156, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0182, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.11617975682020187, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0209, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.1574973613023758, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0222, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.157145157456398, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.0199, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.14698748290538788, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0203, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.12345635890960693, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0212, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.1370248794555664, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0194, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.1332298070192337, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0219, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.12525111436843872, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.0219, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.14905311167240143, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0211, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12063094973564148, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.0262, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.10361988842487335, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0195, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.09919667989015579, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0201, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.11141914129257202, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.0223, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.17785966396331787, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0201, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.126350998878479, + "learning_rate": 2.392710390356112e-05, + "loss": 0.0201, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.12107345461845398, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0213, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.11466097086668015, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0202, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.15397872030735016, + "learning_rate": 2.391713512040237e-05, + "loss": 0.022, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.14412778615951538, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0223, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.10495682060718536, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0194, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.13021183013916016, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0199, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.13993357121944427, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0182, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.128571018576622, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.0181, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.11090676486492157, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0204, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.10988310724496841, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0203, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.11705116927623749, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0214, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.1277252584695816, + "learning_rate": 2.388719222953459e-05, + "loss": 0.0225, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1561414748430252, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0221, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.15278629958629608, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.02, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14865094423294067, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0213, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.1127776950597763, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0211, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.15085943043231964, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0212, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14555132389068604, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0216, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.14708220958709717, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0209, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.1093590185046196, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0199, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.1303861141204834, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0242, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.11462346464395523, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0197, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.12198365479707718, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0206, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.11193065345287323, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0201, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.13254034519195557, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0181, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.13355755805969238, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0204, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12551282346248627, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0212, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1441735178232193, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0217, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.11511780321598053, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0218, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.10427968204021454, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0222, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.11364301294088364, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0166, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.1401003897190094, + "learning_rate": 2.38204570134882e-05, + "loss": 0.0215, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.11645112931728363, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0198, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1340353935956955, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0212, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.0982050895690918, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0208, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.16713382303714752, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0185, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.13803045451641083, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.021, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.15552546083927155, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0233, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.10836438834667206, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0211, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.1831241250038147, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0198, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.12369874119758606, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0193, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.10911241918802261, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0221, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.11441849172115326, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0236, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.11105310916900635, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0206, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.13617953658103943, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0189, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.11214766651391983, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0192, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.12536488473415375, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0206, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.10356712341308594, + "learning_rate": 2.376687582517292e-05, + "loss": 0.019, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.13332195580005646, + "learning_rate": 2.376352133090187e-05, + "loss": 0.0216, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.13996818661689758, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0216, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.14305290579795837, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0204, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.10494769364595413, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0202, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.1461218297481537, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.0197, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.12469783425331116, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0207, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.13164709508419037, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.0207, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.1553804576396942, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0228, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1445176750421524, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0217, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12673451006412506, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0164, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.17502479255199432, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0208, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.11328406631946564, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0196, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12234057486057281, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0193, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.1333971917629242, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0214, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.11227905005216599, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0183, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.1287100911140442, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0198, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.10215240716934204, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0194, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.09786848723888397, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0202, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.11103200167417526, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0219, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.1491294801235199, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0202, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.1364760547876358, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.0198, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.11583423614501953, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0206, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.13896186649799347, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0183, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.0970211774110794, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.019, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.10326618701219559, + "learning_rate": 2.368281439559657e-05, + "loss": 0.022, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.11581890285015106, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0234, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10186298936605453, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0205, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11655393242835999, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0193, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.11990638077259064, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0217, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.10573728382587433, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0177, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.09257248789072037, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.0216, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.11310117691755295, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0188, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.12632012367248535, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0217, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1535339206457138, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0209, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.13133156299591064, + "learning_rate": 2.364907420469738e-05, + "loss": 0.0207, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11225332319736481, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0217, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.10744500905275345, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0212, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.10509403049945831, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0208, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.16215650737285614, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0199, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1208849623799324, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0196, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.14882729947566986, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0229, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.18775896728038788, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0206, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.14199234545230865, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0201, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.1510499119758606, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.021, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.16389605402946472, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0197, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.13032132387161255, + "learning_rate": 2.361188413960337e-05, + "loss": 0.0211, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.13895465433597565, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0222, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.10914628207683563, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0206, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.12215261161327362, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0186, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12954379618167877, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.0193, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.12779726088047028, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.02, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.11825352907180786, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0214, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11001764982938766, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0211, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.10464788973331451, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.02, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.0892137810587883, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0198, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.09811265021562576, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0191, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.14199955761432648, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0196, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.13846102356910706, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0216, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.12102426588535309, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0214, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.12360251694917679, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0199, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.12587665021419525, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.15562282502651215, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.0214, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.11366713047027588, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0192, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.15177540481090546, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0204, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.08908293396234512, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0186, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.1349143236875534, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0197, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.09388674795627594, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0205, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.11973906308412552, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0198, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.1290522664785385, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.02, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.12087979167699814, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0207, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.10237318277359009, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0189, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.10426577180624008, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0204, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13028934597969055, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0197, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.10696206241846085, + "learning_rate": 2.351686222335574e-05, + "loss": 0.0206, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.11024264991283417, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0189, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.11294405907392502, + "learning_rate": 2.35100554725415e-05, + "loss": 0.0228, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.1188240721821785, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0204, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.17485342919826508, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0202, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12582673132419586, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0209, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.18180538713932037, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0202, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.1459868997335434, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0189, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.09818759560585022, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0195, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.1473611295223236, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0253, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.09137327969074249, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0209, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.14026956260204315, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0202, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.1511334776878357, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0182, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.15016451478004456, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0204, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.14407901465892792, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.0205, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.1332983374595642, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0192, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.14312349259853363, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.0195, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.11917808651924133, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0194, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.17896629869937897, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0203, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.12084568291902542, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0199, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.14310620725154877, + "learning_rate": 2.344867863274048e-05, + "loss": 0.0201, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1168353408575058, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0225, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.14718633890151978, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0204, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.14102597534656525, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0197, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.09543738514184952, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0218, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1074792668223381, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0198, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.11555791646242142, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0196, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.12645067274570465, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0191, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.11634501069784164, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.02, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.09201087802648544, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0204, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.1155405044555664, + "learning_rate": 2.341449046471072e-05, + "loss": 0.0215, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.09871169179677963, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0216, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.11332481354475021, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0179, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.12474322319030762, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.0177, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.11339092999696732, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0221, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.1200975552201271, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.0209, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.14456698298454285, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0211, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.1371302306652069, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0165, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.11727641522884369, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0232, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.11336934566497803, + "learning_rate": 2.338366646946928e-05, + "loss": 0.0202, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.10325363278388977, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0209, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.11268003284931183, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0187, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10260385274887085, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.0195, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.10920321941375732, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0225, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.09120934456586838, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0194, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11567433178424835, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0197, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.10144805908203125, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0206, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1246950775384903, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0215, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.100107841193676, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0197, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.15822021663188934, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0202, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13518433272838593, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.02, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12268074601888657, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0196, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.12333078682422638, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0184, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.13588106632232666, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0218, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.12120439857244492, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0197, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.16428445279598236, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0198, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.11615289747714996, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.0182, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.1432264745235443, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0207, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.11717039346694946, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0198, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.10115417838096619, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0193, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.11894490569829941, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0198, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.11940047889947891, + "learning_rate": 2.330810220141532e-05, + "loss": 0.021, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.10992413759231567, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0195, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.10544339567422867, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0182, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.11725642532110214, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.0198, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.09130900353193283, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0209, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.09033828228712082, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0177, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.11721840500831604, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0207, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.12251118570566177, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0215, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.1496249884366989, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0219, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.14808881282806396, + "learning_rate": 2.327710137160376e-05, + "loss": 0.0212, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.145681232213974, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.019, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.12777844071388245, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0173, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.1110587939620018, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0198, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.11348012834787369, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0207, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.11685867607593536, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.02, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.17935989797115326, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0211, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.12398742884397507, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0191, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.10836582630872726, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0196, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.12038073688745499, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0207, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.12211567908525467, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0206, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.1702142357826233, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0227, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12037540227174759, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0191, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.15964926779270172, + "learning_rate": 2.323223255665661e-05, + "loss": 0.018, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.12493535876274109, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0212, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.1359032839536667, + "learning_rate": 2.32253202698687e-05, + "loss": 0.0219, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.18753379583358765, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0228, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.1336824893951416, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0201, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.09901099652051926, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0221, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.12925978004932404, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0206, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.1278320550918579, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.0178, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.12977559864521027, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0176, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10098735243082047, + "learning_rate": 2.320110760314603e-05, + "loss": 0.023, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.12594632804393768, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.0209, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.1140967383980751, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0181, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.1287144273519516, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0195, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.12329071760177612, + "learning_rate": 2.318725808577064e-05, + "loss": 0.0198, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.1339070051908493, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10581240057945251, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0265, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.13716311752796173, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0213, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.14610108733177185, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0195, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.128855898976326, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0185, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.0950121283531189, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0223, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.14444753527641296, + "learning_rate": 2.316299750241968e-05, + "loss": 0.017, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.12585613131523132, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0177, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.1290416419506073, + "learning_rate": 2.315606032389525e-05, + "loss": 0.02, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.10740494728088379, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0191, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11578185111284256, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0211, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.11840540170669556, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0209, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.09531640261411667, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0202, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.10436122864484787, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0188, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.14102807641029358, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0205, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14070506393909454, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0186, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.09883972257375717, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0196, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11524099856615067, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0187, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.11240902543067932, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0176, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.08062270283699036, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.019, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.13965697586536407, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0202, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.11376950889825821, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0218, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.13540907204151154, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0188, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.10888075083494186, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0216, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.10280206799507141, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0178, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.11512117087841034, + "learning_rate": 2.309699448641602e-05, + "loss": 0.02, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.09106973558664322, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0182, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10531709343194962, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0204, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.159554585814476, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0173, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.12161346524953842, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0196, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.08224594593048096, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0175, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.12767191231250763, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0234, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.12864384055137634, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0224, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.10232329368591309, + "learning_rate": 2.3069137227667e-05, + "loss": 0.0208, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.09609836339950562, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0183, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.1301981806755066, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0202, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.1258714497089386, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0217, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.12543867528438568, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13326771557331085, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.022, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.11770334839820862, + "learning_rate": 2.304821853869955e-05, + "loss": 0.019, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.11729256063699722, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0187, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.11887779086828232, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0224, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.10921293497085571, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0192, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.1346455216407776, + "learning_rate": 2.303426051834282e-05, + "loss": 0.022, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.1177918091416359, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0209, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.12912803888320923, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0191, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.12571996450424194, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0211, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.13959148526191711, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0198, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.1404334306716919, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0211, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.1111592948436737, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0213, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.10322082787752151, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0203, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.11235317587852478, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0187, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.13599546253681183, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0205, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11067865788936615, + "learning_rate": 2.299932279415038e-05, + "loss": 0.02, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.14130094647407532, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0204, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.14370687305927277, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0199, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.13825304806232452, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0197, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.11885961890220642, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.02, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.12376029789447784, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0199, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.10103051364421844, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0181, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.10931621491909027, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0198, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.11835317313671112, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0202, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.08036374300718307, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0189, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.08135218918323517, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0196, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.12611854076385498, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0211, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.09217566251754761, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0181, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.10258878767490387, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0194, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.1165921688079834, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0196, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.09290144592523575, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.0205, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.11575523763895035, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0204, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13866478204727173, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0178, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.13036414980888367, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0191, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11062757670879364, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0193, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.09973529726266861, + "learning_rate": 2.292926536293724e-05, + "loss": 0.02, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.13707594573497772, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0186, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11635670065879822, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0185, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.13926011323928833, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0192, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.13966591656208038, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0194, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.09878037869930267, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0191, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.11935862898826599, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0193, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.15291833877563477, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0189, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.12976837158203125, + "learning_rate": 2.290117482755701e-05, + "loss": 0.017, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.13367557525634766, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0178, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12121888250112534, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0191, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.13135644793510437, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0226, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.1142699271440506, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.017, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.09670589119195938, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0199, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.12751951813697815, + "learning_rate": 2.288008171679078e-05, + "loss": 0.018, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.14463315904140472, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0184, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.11006469279527664, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0198, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.15361396968364716, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0186, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.20782725512981415, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0208, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.15324300527572632, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.018, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.17235179245471954, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0218, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.12993113696575165, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0198, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.1496446281671524, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0203, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.16271376609802246, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.0215, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11106697469949722, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0206, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.10508760809898376, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0218, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.11433881521224976, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0203, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.12877102196216583, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0203, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.13301214575767517, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0225, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.15351711213588715, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0213, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.1114102303981781, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0212, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.09978143125772476, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0198, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10524465143680573, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0183, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.09780002385377884, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0185, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.1361900269985199, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0218, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.1329333484172821, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0191, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.10501913726329803, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.019, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.1329827904701233, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0198, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.10849577933549881, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0213, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.14308017492294312, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.0199, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.12313494831323624, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0193, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.10307908058166504, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.0215, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10591195523738861, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0178, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.1411430686712265, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0237, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13116225600242615, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0203, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.12032114714384079, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0201, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.12519164383411407, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0177, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.10340812057256699, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0194, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.10876207798719406, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0207, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11245796829462051, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0207, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11444602906703949, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0171, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12232744693756104, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0196, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.10049490630626678, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0179, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.09973384439945221, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.022, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.11037951707839966, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0197, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.14473563432693481, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0207, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.1138916090130806, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0198, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.12471994757652283, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0182, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.12418543547391891, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0195, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.12200118601322174, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.0217, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.13389603793621063, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0194, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10176926106214523, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.0189, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.10195910185575485, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0202, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.13289214670658112, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.018, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.10307228565216064, + "learning_rate": 2.270347347405331e-05, + "loss": 0.0183, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.149918794631958, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0172, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.1182866096496582, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0173, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.1035924181342125, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0194, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.108797088265419, + "learning_rate": 2.268928118647676e-05, + "loss": 0.0174, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.11457139253616333, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0208, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.10256518423557281, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0188, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13053470849990845, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0209, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.11632310599088669, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0187, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.12769649922847748, + "learning_rate": 2.267152769015067e-05, + "loss": 0.0206, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.13292407989501953, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0191, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.09822595119476318, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0189, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.13729946315288544, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0229, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.15705017745494843, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0203, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.12355367839336395, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0201, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.1278308629989624, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0205, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.09874208271503448, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0218, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.1316995620727539, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0212, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.10361800342798233, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0209, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.09672107547521591, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0205, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.12609392404556274, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0205, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.1360992044210434, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0188, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.11627763509750366, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0206, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.11733578145503998, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.02, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.1705910563468933, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0174, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.11704754829406738, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0199, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.13247141242027283, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0224, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.10916013270616531, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0202, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.11092681437730789, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.0186, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.12811405956745148, + "learning_rate": 2.260036839187291e-05, + "loss": 0.02, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.11893392354249954, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0194, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.16553092002868652, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0189, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.12431131303310394, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.02, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10850611329078674, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0222, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.12647055089473724, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0193, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.10432784259319305, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0196, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.1159503236413002, + "learning_rate": 2.257540798155951e-05, + "loss": 0.0192, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.10592516511678696, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0219, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.1318269520998001, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0195, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.11932560801506042, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0194, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.16499531269073486, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.02, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.1334158480167389, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.0217, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.14781776070594788, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0206, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.14816460013389587, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0206, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.11319883167743683, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0196, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.12413972616195679, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0206, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.1277540624141693, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0211, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.12178509682416916, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0192, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.1210661232471466, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0232, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.11300696432590485, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0182, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.15876832604408264, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0194, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.1387646645307541, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0202, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.12502169609069824, + "learning_rate": 2.251824990920905e-05, + "loss": 0.02, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.12365051358938217, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0197, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.12610606849193573, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0194, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.13096576929092407, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0208, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.13002553582191467, + "learning_rate": 2.250393750922939e-05, + "loss": 0.02, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.10048994421958923, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0193, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.11463778465986252, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0179, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10637608915567398, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0182, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.11959703266620636, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0183, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.11852603405714035, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12075617164373398, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0186, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.09626025706529617, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0195, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12773670256137848, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.019, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13364142179489136, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.019, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.14562146365642548, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0207, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.11752697080373764, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.13970085978507996, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0219, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.12456551194190979, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0203, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.36701101064682007, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0191, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.17254851758480072, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0193, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.1397981345653534, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0205, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.1273665875196457, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0227, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.08652572333812714, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0194, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.118954136967659, + "learning_rate": 2.243582938893974e-05, + "loss": 0.019, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.12105695903301239, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0173, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.10059279948472977, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0189, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.1476186215877533, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0192, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.11198431998491287, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0212, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.14829197525978088, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0216, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.14479608833789825, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0225, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.16737256944179535, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0202, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.12101127207279205, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0198, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.09408389776945114, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.019, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.10649813711643219, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0191, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.1328127533197403, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.02, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1135721281170845, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0181, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.14464536309242249, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0167, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.15146523714065552, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0216, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11212945729494095, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0189, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.11722676455974579, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0185, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10031717270612717, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0181, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.14852827787399292, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.02, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12114712595939636, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0207, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.12325385212898254, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0185, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.12915199995040894, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0188, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.10423776507377625, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0195, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.13398106396198273, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0182, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.1426725685596466, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0214, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.12515439093112946, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0184, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.15888792276382446, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0183, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.11037078499794006, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0199, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.10706029832363129, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.0195, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.11116144061088562, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0202, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.12716875970363617, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0182, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.1446460634469986, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0172, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.09947994351387024, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0205, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.09251125156879425, + "learning_rate": 2.231705291760663e-05, + "loss": 0.018, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.08440345525741577, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0181, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12813445925712585, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0206, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.1017315611243248, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0192, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.13247326016426086, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0212, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.14845387637615204, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0194, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.13824601471424103, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0224, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.1151830404996872, + "learning_rate": 2.229177978101565e-05, + "loss": 0.0176, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.12420724332332611, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0179, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.13085328042507172, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0179, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.12436439841985703, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.11873050034046173, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0197, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.1338828206062317, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0212, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.1106492280960083, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0179, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.16542823612689972, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0202, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.08829429000616074, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0176, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.10792568325996399, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0186, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.1279839128255844, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0221, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.10742732882499695, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0217, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.10220900177955627, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0201, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.1660427749156952, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.019, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.13747647404670715, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0225, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.13351675868034363, + "learning_rate": 2.223753180973339e-05, + "loss": 0.021, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.14151513576507568, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.0195, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.12823058664798737, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0189, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.12779751420021057, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0177, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.2243196815252304, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0183, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.12365520745515823, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0188, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.10981941968202591, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0191, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.11569637805223465, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0195, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.15785852074623108, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0222, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11610189825296402, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0172, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.1124223843216896, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0184, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.14634257555007935, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.0196, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.13306891918182373, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0204, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.08198657631874084, + "learning_rate": 2.219041683306619e-05, + "loss": 0.0177, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.09529033303260803, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0166, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.11170824617147446, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0195, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13825464248657227, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0191, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.11621235311031342, + "learning_rate": 2.217590132288247e-05, + "loss": 0.0202, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11921709030866623, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.019, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12485305964946747, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0205, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.1244436725974083, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0204, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11240405589342117, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0197, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.1087997704744339, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0183, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.11403695493936539, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0172, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.12764202058315277, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0161, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.17418256402015686, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0171, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.1522650271654129, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0197, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.16773882508277893, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0214, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.14681261777877808, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0208, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.14842623472213745, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.021, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.10614728182554245, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0182, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.15053948760032654, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0209, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.1381484866142273, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.0176, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.13161014020442963, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0174, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.09695997089147568, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0202, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.08903522044420242, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0171, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11456362903118134, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0183, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.1260307878255844, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.02, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.11732659488916397, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0197, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.126072496175766, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0196, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.11586430668830872, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.02, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.10504680871963501, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0193, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.09652367234230042, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0186, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10370496660470963, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0176, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.1192321851849556, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0178, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.10527302324771881, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0182, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10162831097841263, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0191, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.16128556430339813, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0203, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.11897512525320053, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0204, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.14880484342575073, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0205, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.11369120329618454, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.0193, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.11618658900260925, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0176, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.10052158683538437, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.0207, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1014244481921196, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0205, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.15118759870529175, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0215, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.09245041012763977, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0176, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.13509149849414825, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0224, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.12747371196746826, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0185, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11445175856351852, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0208, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11949758231639862, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0193, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.12797918915748596, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0182, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0185, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.11956185847520828, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0189, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.11755053699016571, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0184, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.1273607611656189, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0205, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.11855774372816086, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0188, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.09031623601913452, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0183, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12208469212055206, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0199, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.1083436831831932, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0193, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12228160351514816, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0182, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.1007765457034111, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0195, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.11114434152841568, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0173, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.1096438392996788, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0199, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.124867744743824, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0219, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.10032925009727478, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.0194, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.13164062798023224, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0193, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.1376691609621048, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0187, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.11254525184631348, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0186, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.12835443019866943, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.15313562750816345, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0219, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.16989216208457947, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0172, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.11433102935552597, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.019, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09817333519458771, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0222, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.09500890970230103, + "learning_rate": 2.193514973491427e-05, + "loss": 0.0176, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.14314892888069153, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0199, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.11764178425073624, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0169, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.12384455651044846, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0193, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.11438048630952835, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0174, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.13921929895877838, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0182, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.16880372166633606, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0174, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.1554516702890396, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0194, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.15355393290519714, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0212, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.1179167702794075, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0195, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.13388752937316895, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0208, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.10997489839792252, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0181, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.11783649772405624, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0189, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13451549410820007, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0186, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.1339322179555893, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0179, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.09972941130399704, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0214, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.10101358592510223, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0176, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.09771978110074997, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0177, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.14234617352485657, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0184, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.10462982952594757, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0197, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.09772231429815292, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0226, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.12018870562314987, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0164, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.10980863869190216, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0199, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.11285397410392761, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0172, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.1067347601056099, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0183, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.1046559289097786, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.0198, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.11118492484092712, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0179, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.1001342311501503, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0166, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.1053142324090004, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0188, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.16253869235515594, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.018, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.10042452812194824, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0173, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.11815430223941803, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0181, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.07585619390010834, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0195, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.17831926047801971, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0183, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.09953836351633072, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0194, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.12129344046115875, + "learning_rate": 2.18065424687869e-05, + "loss": 0.0188, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.09982063621282578, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0196, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.0906965360045433, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0172, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.1256277859210968, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.019, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.10290617495775223, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0171, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.17007306218147278, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0178, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.0779242068529129, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0192, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.12276264280080795, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0233, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.13211409747600555, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0184, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11407056450843811, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0211, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.13331526517868042, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0193, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.1603303849697113, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0192, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.10634022206068039, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0174, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.1658177524805069, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0207, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.1432579606771469, + "learning_rate": 2.175492159090619e-05, + "loss": 0.02, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.11264102905988693, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0183, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.12988954782485962, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.0199, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.1687532663345337, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0175, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12261682003736496, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0148, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.13096439838409424, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0175, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.13871102035045624, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0198, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.12536168098449707, + "learning_rate": 2.172907339903944e-05, + "loss": 0.018, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.12913931906223297, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0194, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.1255621761083603, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0172, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.12288706004619598, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0206, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.113137386739254, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0201, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.11107201129198074, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.14609375596046448, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.02, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.15227735042572021, + "learning_rate": 2.170320016677318e-05, + "loss": 0.016, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.12586566805839539, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0194, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.13495711982250214, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0187, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.11436781287193298, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.018, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.12271122634410858, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0193, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.16221405565738678, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0195, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.08645932376384735, + "learning_rate": 2.168100325341058e-05, + "loss": 0.0157, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.11750698834657669, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0186, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.1191735714673996, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0178, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.10067173838615417, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0187, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.09761539846658707, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0169, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.13806267082691193, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0217, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.10618548095226288, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0165, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.15135695040225983, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0199, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10563432425260544, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0176, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.12757790088653564, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0183, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.11258719861507416, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.0162, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.15290242433547974, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0192, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.1350955218076706, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0188, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.09494426846504211, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0187, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12816165387630463, + "learning_rate": 2.162913952762737e-05, + "loss": 0.019, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.1237507089972496, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0187, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.12867869436740875, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0202, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11882596462965012, + "learning_rate": 2.161801300808089e-05, + "loss": 0.019, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.10291648656129837, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0187, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.09269602596759796, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.019, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.08865118771791458, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.0188, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10444393008947372, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0182, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.11461203545331955, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.0219, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.13728564977645874, + "learning_rate": 2.159574640654742e-05, + "loss": 0.0223, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.1394069343805313, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0172, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.15394151210784912, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0182, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.10422483086585999, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0175, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.10996434837579727, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0182, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.1026761382818222, + "learning_rate": 2.157717712784381e-05, + "loss": 0.018, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.1373908817768097, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0226, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.10155291855335236, + "learning_rate": 2.156974591793542e-05, + "loss": 0.0179, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.14904215931892395, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0192, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.1131916344165802, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0175, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.09735766798257828, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0182, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.13132493197917938, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0193, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.12934266030788422, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0184, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.1196565330028534, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0186, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.12375814467668533, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.0199, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.0892246812582016, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0178, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.0932748094201088, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0192, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11503782123327255, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0187, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.08250989019870758, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0183, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.101202592253685, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0183, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1226576566696167, + "learning_rate": 2.152139450890464e-05, + "loss": 0.022, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.1167614683508873, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0176, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.12396091967821121, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0186, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.13461633026599884, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.0169, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.10792742669582367, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0189, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11649876832962036, + "learning_rate": 2.150277549265773e-05, + "loss": 0.02, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.1314917951822281, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0157, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.14143578708171844, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0197, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10042720288038254, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0177, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.12019779533147812, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0164, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.14096254110336304, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0223, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.10627058893442154, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0171, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.11092417687177658, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.019, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.12211407721042633, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0199, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11551571637392044, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0197, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.10474767535924911, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.018, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.12364303320646286, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0178, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13243450224399567, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0197, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10594736039638519, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0179, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.11143538355827332, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0196, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.10518378764390945, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0197, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12268701940774918, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0176, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.09408178925514221, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0205, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.12529458105564117, + "learning_rate": 2.143564506261267e-05, + "loss": 0.0184, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.15819092094898224, + "learning_rate": 2.143191093594534e-05, + "loss": 0.021, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.15818199515342712, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0182, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.13374921679496765, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0177, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.10046645253896713, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0197, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.11422070860862732, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0185, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.1138019785284996, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0182, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.09617039561271667, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0178, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.1452113538980484, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0222, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.10409227013587952, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0196, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.11277221143245697, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0175, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.12325838208198547, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0201, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12721778452396393, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0174, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.14016005396842957, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.017, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.11149191111326218, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0168, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.13445520401000977, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0194, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10086208581924438, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0197, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.115168996155262, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0181, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.11483300477266312, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.0183, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.1196938082575798, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0172, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.12219196557998657, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0177, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.11318044364452362, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.0204, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.09219531714916229, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0181, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.09911682456731796, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0192, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.11822745203971863, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0192, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.11101230978965759, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0165, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.09978491812944412, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0203, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.09874484688043594, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0199, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.08682837337255478, + "learning_rate": 2.133090580749213e-05, + "loss": 0.0177, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.14352355897426605, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0189, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.11133426427841187, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0199, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.10667043924331665, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0183, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.14207594096660614, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0218, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.12165568023920059, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0167, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.09311263263225555, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0183, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10423581302165985, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0151, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.12640471756458282, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0214, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.11349574476480484, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0171, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.14359329640865326, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0182, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.10977838933467865, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0181, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.12443531304597855, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0168, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.12441008538007736, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0186, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.14103129506111145, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0197, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10460726171731949, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0178, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.12526848912239075, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0188, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.13363681733608246, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0183, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.13106584548950195, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0181, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.1353096216917038, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0202, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.11889169365167618, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0198, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.12628430128097534, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0178, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.13260240852832794, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0182, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12357809394598007, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0175, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.13913387060165405, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0189, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.12303008884191513, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0187, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.10231908410787582, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.017, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.20065239071846008, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0172, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.11311347782611847, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.019, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.131951242685318, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0182, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.15945149958133698, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0202, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09272884577512741, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0189, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.0937163308262825, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0208, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.10963460803031921, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0197, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.1618727743625641, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0195, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.0919048935174942, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0177, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.12060586363077164, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0209, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.13121892511844635, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0181, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.11795026063919067, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0186, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.13060589134693146, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0226, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.08142203092575073, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0181, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.12634120881557465, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0178, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.1448773890733719, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0179, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1230989471077919, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0188, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.11944684386253357, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0212, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.1445692479610443, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0184, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.0831015482544899, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0171, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.1456967294216156, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0189, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.11207464337348938, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0185, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.16137553751468658, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0182, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.09820980578660965, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0192, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.11823676526546478, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0183, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.12873874604701996, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0181, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.11243417114019394, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0188, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.11120739579200745, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0222, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.10704528540372849, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0178, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.12063440680503845, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0174, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.13275623321533203, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0171, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.10400331765413284, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0201, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.10485411435365677, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0187, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.12517869472503662, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0187, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.11703156679868698, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.0178, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.11982452869415283, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0186, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11858765780925751, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0182, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.10404367744922638, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0173, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.09697552025318146, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0188, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.12046534568071365, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0173, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.0896986722946167, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0188, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.10993069410324097, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0213, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.1200791671872139, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0179, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.14664512872695923, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.0202, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.09285914152860641, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0207, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.1080656424164772, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0188, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.140005961060524, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0188, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.11084222793579102, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.02, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.0913061872124672, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0194, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.09459970891475677, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0182, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.09709502756595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0195, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.11101861298084259, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0184, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.1152782216668129, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0178, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.11495838314294815, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0178, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.14250393211841583, + "learning_rate": 2.10258099183954e-05, + "loss": 0.019, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.09898682683706284, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0173, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.13904057443141937, + "learning_rate": 2.101823807728166e-05, + "loss": 0.018, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.12013959139585495, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0163, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.0925723984837532, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0187, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.1284961700439453, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0158, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.12408998608589172, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0206, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1095876470208168, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0172, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.10166801512241364, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0183, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13473083078861237, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0155, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.10492225736379623, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0188, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.10769215226173401, + "learning_rate": 2.098414220174082e-05, + "loss": 0.021, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.09331739693880081, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0176, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.09247364103794098, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0163, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.20951682329177856, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0203, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.16750548779964447, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0196, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.10933154821395874, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0169, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10767697542905807, + "learning_rate": 2.09613911545763e-05, + "loss": 0.0184, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.11565674096345901, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0208, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.11934554576873779, + "learning_rate": 2.095380384728948e-05, + "loss": 0.018, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.12786556780338287, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0169, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.11752485483884811, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0194, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.11586920917034149, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0179, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.08849998563528061, + "learning_rate": 2.0938623809196e-05, + "loss": 0.018, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.12535157799720764, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0176, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.10429015755653381, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0188, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.09504082053899765, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0186, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.12917572259902954, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0182, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.11084037274122238, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0163, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.12090355157852173, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0196, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.11049258708953857, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0194, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.0831870436668396, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0188, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11006250977516174, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.018, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.12107989192008972, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0178, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.10361241549253464, + "learning_rate": 2.089684155251222e-05, + "loss": 0.017, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.13428105413913727, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0198, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.13375632464885712, + "learning_rate": 2.088923894557553e-05, + "loss": 0.018, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.14561115205287933, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0181, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.09850531816482544, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0176, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.11646582931280136, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0162, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.14659325778484344, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0168, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.15971294045448303, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0198, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.11780509352684021, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0156, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11453750729560852, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0179, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.09094218909740448, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0172, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.13252301514148712, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0209, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.10068004578351974, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0159, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.10302501171827316, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0172, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.10893740504980087, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.019, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.12446500360965729, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0177, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1345142424106598, + "learning_rate": 2.083597073454921e-05, + "loss": 0.021, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.12400489300489426, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0194, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.15042100846767426, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0162, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.1559734046459198, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0189, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.10689649730920792, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0163, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.15904614329338074, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0199, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.11539188027381897, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0208, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.14550678431987762, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0191, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.10212639719247818, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0177, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.10318224877119064, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0176, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0196, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.10269734263420105, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0179, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.11070621758699417, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0196, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13299307227134705, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.02, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11430692672729492, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0188, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10718249529600143, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.0204, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.11855681240558624, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.019, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.10754020512104034, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0157, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11011523753404617, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0172, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.11931152641773224, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.0171, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.12425696104764938, + "learning_rate": 2.075972279799841e-05, + "loss": 0.0195, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.10292454063892365, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0177, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.11290059238672256, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0165, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.1453622728586197, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0207, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.12152253836393356, + "learning_rate": 2.07444521407078e-05, + "loss": 0.017, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.134286567568779, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0198, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.11180262267589569, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0196, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.12045837193727493, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.018, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.18509334325790405, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0209, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.17120368778705597, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0197, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.11996874213218689, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0204, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.12391912192106247, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0201, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.1248779147863388, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.016, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.12873846292495728, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0184, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.11072682589292526, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.0182, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.10367089509963989, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0179, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.12854290008544922, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0184, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.16348853707313538, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0169, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.1345866173505783, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0187, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.1139485239982605, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0178, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.1656051129102707, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0177, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.1133129820227623, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0154, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.0865294560790062, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0163, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.1368715763092041, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0218, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.11601361632347107, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0185, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.1401238590478897, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0177, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.10183753073215485, + "learning_rate": 2.066033925513678e-05, + "loss": 0.0192, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.10317160934209824, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0153, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.12317138910293579, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.021, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.10953490436077118, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0191, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.10530596971511841, + "learning_rate": 2.064502355956822e-05, + "loss": 0.0179, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.09054628014564514, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0189, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.12472135573625565, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0239, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.1551167368888855, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0189, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.08141466975212097, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0174, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.13025878369808197, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0185, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.13056428730487823, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0183, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.1187913790345192, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0182, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.11957527697086334, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0173, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.09339714050292969, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0182, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.1584015190601349, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.02, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.13754191994667053, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0193, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.09880727529525757, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0192, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.15097497403621674, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0172, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12028415501117706, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0182, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.10583297163248062, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0183, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.09343970566987991, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0187, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.1436399519443512, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0175, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.10790709406137466, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0184, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.14520230889320374, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0187, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.10826292634010315, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0188, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.1122521162033081, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0197, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.11445961147546768, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.015, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.1473265290260315, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0193, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.1572754830121994, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0186, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.1085842102766037, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0169, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.11982112377882004, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0163, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.12014015763998032, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.0217, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.11440391838550568, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0175, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.1276320368051529, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0203, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12434501200914383, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0186, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.11981593072414398, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0138, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.08389589190483093, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0162, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.11693917959928513, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.0199, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.1033058762550354, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0182, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.17878879606723785, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0181, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.12847506999969482, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0186, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.10514750331640244, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0166, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.1037413701415062, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0182, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.10402073711156845, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0165, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.10330315679311752, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0164, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.14082680642604828, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0205, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.1187078058719635, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0178, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.10590347647666931, + "learning_rate": 2.047995045386578e-05, + "loss": 0.0192, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.11178472638130188, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.019, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.11713241040706635, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0169, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.1520795077085495, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0173, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.11906655877828598, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0163, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.10854434967041016, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0181, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.16785599291324615, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.0212, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10305774211883545, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0179, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.09346644580364227, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0162, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11048682034015656, + "learning_rate": 2.044530211206255e-05, + "loss": 0.0158, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.14503270387649536, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0185, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.12631897628307343, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0182, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.14716260135173798, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0205, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.1364247053861618, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0187, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.09775770455598831, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0179, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.17999885976314545, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0194, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.1381615400314331, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0192, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.10749474912881851, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0189, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.15220865607261658, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.12083570659160614, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0175, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.11410468816757202, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0174, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.11285928636789322, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.0191, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.09808669239282608, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0173, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.07822877913713455, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0179, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.1032382920384407, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0177, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12453432381153107, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0183, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.1397278904914856, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0175, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.1701551228761673, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0176, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.13178907334804535, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0158, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.16403238475322723, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0182, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.13050618767738342, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.019, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09723208099603653, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.016, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.11937404423952103, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0194, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.10595990717411041, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0179, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.10598801076412201, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0172, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.12562280893325806, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0192, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.1062425896525383, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0167, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.19852802157402039, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0186, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10752324014902115, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0171, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.11313100904226303, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0191, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.09881386160850525, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0189, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.11926184594631195, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0176, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.11304130405187607, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0171, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.159687340259552, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0209, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.12324368953704834, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0184, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.10718217492103577, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0178, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12370229512453079, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0181, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.11843078583478928, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.017, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.11866020411252975, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0173, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.10999849438667297, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0217, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.10843008756637573, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.0163, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.13494807481765747, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0202, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.13634233176708221, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.0203, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.09910089522600174, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0154, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.1105995923280716, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0172, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.11397486180067062, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0171, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10141155868768692, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0187, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.09431210905313492, + "learning_rate": 2.025995044916193e-05, + "loss": 0.0163, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.09627058357000351, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0188, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.10498029738664627, + "learning_rate": 2.025220726516947e-05, + "loss": 0.0197, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.12692295014858246, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0163, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.11992908269166946, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0188, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.1448463350534439, + "learning_rate": 2.024058949838742e-05, + "loss": 0.0182, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.10764076560735703, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0191, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.1176103726029396, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0202, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.10393970459699631, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0174, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.12357908487319946, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0188, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11184806376695633, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0171, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09287009388208389, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0164, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.13145606219768524, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0175, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.108634814620018, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0192, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.1643940508365631, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.02, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.1642858386039734, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0195, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.11515326052904129, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0173, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.11438507586717606, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0187, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.1556805521249771, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0193, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.1081680878996849, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0176, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.11440131068229675, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0198, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.13953827321529388, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0163, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.12443852424621582, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0173, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.12884436547756195, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.0184, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.09170050173997879, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0158, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.08685769885778427, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0143, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.12880417704582214, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0172, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.09957600384950638, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0189, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.11359020322561264, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0191, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.10620398819446564, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0183, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.11106064170598984, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0175, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.14327868819236755, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0156, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.09743347764015198, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0165, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.11041118204593658, + "learning_rate": 2.013198534613597e-05, + "loss": 0.019, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.13757234811782837, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.017, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.10167638957500458, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0172, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.1386769413948059, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0157, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.13493874669075012, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0184, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.1238163560628891, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0178, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12875531613826752, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0176, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11199075728654861, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0179, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12402849644422531, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0197, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.12012362480163574, + "learning_rate": 2.009701166604228e-05, + "loss": 0.0203, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.1319619119167328, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0176, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.11875088512897491, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.017, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.14083963632583618, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0192, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1157609224319458, + "learning_rate": 2.008145773123833e-05, + "loss": 0.017, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1447380781173706, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0192, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.08668482303619385, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.0182, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.11417029052972794, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0162, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.12431302666664124, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0176, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.09466125816106796, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0174, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.13191643357276917, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0162, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.16682866215705872, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0167, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.09262735396623611, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0174, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10246162116527557, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0157, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.14533615112304688, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0193, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10985267907381058, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0185, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.1364467293024063, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0174, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.12792983651161194, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.019, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.13055621087551117, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0175, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.1272381991147995, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0191, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.1523309350013733, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.0192, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.12438428401947021, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0183, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.13417088985443115, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0192, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.10955093055963516, + "learning_rate": 2.000749254391884e-05, + "loss": 0.019, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.1140032708644867, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.0155, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.13144028186798096, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.0177, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.10048601031303406, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0169, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.1272922307252884, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0167, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.10301118344068527, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0166, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.10341407358646393, + "learning_rate": 1.998410649877165e-05, + "loss": 0.018, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.08479570597410202, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0193, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.10605745017528534, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0169, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.09806226938962936, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0175, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.12049946188926697, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0187, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.13633446395397186, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0167, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.12669390439987183, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.02, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.1395879089832306, + "learning_rate": 1.99568055617237e-05, + "loss": 0.02, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.1145915761590004, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.019, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12087064981460571, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0194, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.11252355575561523, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0156, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.10857690125703812, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0197, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.17054887115955353, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0172, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.12123709172010422, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0184, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.08856352418661118, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0178, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.09841718524694443, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0198, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.1236206591129303, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0171, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11047739535570145, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0198, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.11695606261491776, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0162, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.12013910710811615, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0171, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.08011548966169357, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0176, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.13469302654266357, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0177, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.12169095873832703, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.0192, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.1165102869272232, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0165, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.16436143219470978, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0164, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.11628631502389908, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0174, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.13279633224010468, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0164, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11530247330665588, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0199, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13026569783687592, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0193, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.12264525890350342, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0178, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.09320709109306335, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0157, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.11837475001811981, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.016, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.12293318659067154, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.0182, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.09163888543844223, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0197, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.1171371266245842, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0162, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.1096082329750061, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0198, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09364283084869385, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0164, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.08231621980667114, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0169, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.1273244321346283, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.0167, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.15531282126903534, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0194, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.13366153836250305, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0186, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.1072721853852272, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0173, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.11176268011331558, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0182, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.10074266046285629, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0167, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.11638782918453217, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0194, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.1109209880232811, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0197, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.1511884480714798, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0202, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.09859053790569305, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0185, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.12002220004796982, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0177, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.11930137127637863, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0159, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.10564088076353073, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0176, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.10428094118833542, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0197, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.1335987150669098, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0188, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.12865577638149261, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0179, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.14246998727321625, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0181, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.1250566691160202, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0188, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.09627831727266312, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0153, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.09008506685495377, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0185, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.12598916888237, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0169, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.11335116624832153, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0165, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.10588988661766052, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0171, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.10731689631938934, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.0174, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.11184538155794144, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0154, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.10870286822319031, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0163, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.11692065745592117, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0178, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.10036200284957886, + "learning_rate": 1.972989772246684e-05, + "loss": 0.016, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.12077154219150543, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0178, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.1272457093000412, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0188, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.08074570447206497, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0159, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.11381031572818756, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.0155, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.09836770594120026, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0161, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.09236133843660355, + "learning_rate": 1.970635454034925e-05, + "loss": 0.0183, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.1167517825961113, + "learning_rate": 1.970242942396647e-05, + "loss": 0.018, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.10953414440155029, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.015, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.12463351339101791, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0192, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.1285952925682068, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0172, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.10794857144355774, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0167, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.11815376579761505, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0176, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.10251157730817795, + "learning_rate": 1.967887123661664e-05, + "loss": 0.018, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.16941046714782715, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0182, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.14373992383480072, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.0177, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.11837214231491089, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0169, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.10599380731582642, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0167, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.11006144434213638, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0207, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.11699084937572479, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0164, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.10131341218948364, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0177, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.09975451976060867, + "learning_rate": 1.964744043284577e-05, + "loss": 0.017, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.08056963980197906, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0171, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.16518834233283997, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0186, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.10148414224386215, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0181, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.14867499470710754, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0183, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.16952896118164062, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0199, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13692231476306915, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0198, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.15133927762508392, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0184, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11434897035360336, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0189, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.0917418971657753, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0167, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12218283116817474, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0168, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.12193279713392258, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0173, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.09389539062976837, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0182, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.10455376654863358, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0169, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.11151516437530518, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0188, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.14512453973293304, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0194, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.10494859516620636, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0169, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.15839508175849915, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0177, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.12615326046943665, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0172, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.1135321632027626, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0166, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.12872013449668884, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0173, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.18429434299468994, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0165, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.13249747455120087, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0183, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.10603360831737518, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0145, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.10620120912790298, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0155, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.10986930131912231, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0163, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09298109263181686, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0179, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09945658594369888, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0178, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.08814677596092224, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0186, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.09042713046073914, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0173, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.11592321842908859, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0191, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.1026391014456749, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.018, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.09040834754705429, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0164, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.11391181498765945, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0177, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11113253980875015, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.0166, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.10178712755441666, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0167, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.13696520030498505, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0171, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.10290519148111343, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0164, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13472187519073486, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0173, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.08172356337308884, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0187, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.12297531962394714, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0166, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.13925355672836304, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0193, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.15394000709056854, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0169, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.09050899744033813, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.019, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.11854210495948792, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0181, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.10207716375589371, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0154, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.10006722062826157, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0161, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.10565274208784103, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0183, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.11656428128480911, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0166, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.12020296603441238, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0181, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.0964578241109848, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0174, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.0888240784406662, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0158, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.09271078556776047, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0187, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.10407985746860504, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0152, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1329568326473236, + "learning_rate": 1.94346993525268e-05, + "loss": 0.0166, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.13284213840961456, + "learning_rate": 1.943075029574398e-05, + "loss": 0.018, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.14158464968204498, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0194, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.11495806276798248, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0157, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.12886112928390503, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0174, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08954162895679474, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0158, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.09367296099662781, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0171, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.11188989877700806, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0162, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.10758768022060394, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0154, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.11783428490161896, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0182, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.18594273924827576, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0175, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.08010479807853699, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0167, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13949649035930634, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0175, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.10709669440984726, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0179, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.1675017923116684, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0176, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12764538824558258, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0165, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11073391884565353, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0169, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.10426445305347443, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0171, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.09000374376773834, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0161, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.10287991911172867, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0171, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.12322332710027695, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0178, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.11188322305679321, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0166, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.14056499302387238, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0196, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.08969348669052124, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0169, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.10100627690553665, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0159, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.1181580200791359, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0185, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.11281560361385345, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0171, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.11793895065784454, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0187, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.14534896612167358, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0177, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.10177207738161087, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0182, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.1335727721452713, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.017, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.14608687162399292, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0197, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.09763725847005844, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0184, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.09486788511276245, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0177, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.12235541641712189, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.018, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.13838450610637665, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0174, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.09837213158607483, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.0174, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.11713139712810516, + "learning_rate": 1.928836248385875e-05, + "loss": 0.016, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.13015268743038177, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0187, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.13498486578464508, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0181, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.12189928442239761, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0156, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.14399787783622742, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0185, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.10662715882062912, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0164, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.1145375519990921, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0185, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.11677944660186768, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0153, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.10242128372192383, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0168, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.1290024369955063, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0189, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.11820637434720993, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0167, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.12113048881292343, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0187, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.12122460454702377, + "learning_rate": 1.92408053936279e-05, + "loss": 0.0175, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.09413880109786987, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0149, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.11389796435832977, + "learning_rate": 1.923287468901755e-05, + "loss": 0.0185, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13923633098602295, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.0169, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.15947312116622925, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0188, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.14314432442188263, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0182, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.1426798403263092, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0179, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.12179753184318542, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.02, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11758018285036087, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0197, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11349253356456757, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0171, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.13724955916404724, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0181, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.1260363608598709, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0172, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.09541817009449005, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0181, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.0922083631157875, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0177, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.08051180839538574, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0167, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.11803245544433594, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0177, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10018756985664368, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.10512874275445938, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0164, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.17478078603744507, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0169, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.08864858746528625, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0165, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11313372850418091, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0172, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.09432164579629898, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0163, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.09435626119375229, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0188, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.10041314363479614, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0157, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.07982632517814636, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0176, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.09699293971061707, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0177, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.11059583723545074, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.017, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.11384644359350204, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0163, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.11473467200994492, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0165, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.11018316447734833, + "learning_rate": 1.912568530871305e-05, + "loss": 0.0179, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.10781396180391312, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0199, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.11702359467744827, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0165, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.12909658253192902, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0162, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.09839031100273132, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0152, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.1357898861169815, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0198, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.11221814155578613, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.0178, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.11431093513965607, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0176, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.11436092853546143, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0187, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.09838918596506119, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0148, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.11489785462617874, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0168, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.10876795649528503, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0169, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.1032189428806305, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0152, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.10013218969106674, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.11554815620183945, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0182, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.2529490292072296, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0198, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.09215852618217468, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0168, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.12875677645206451, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0185, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.1232401430606842, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0179, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.10334797203540802, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0171, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10524798929691315, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0162, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.11014945805072784, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0176, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.1074177473783493, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0162, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.14043818414211273, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0191, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.10011711716651917, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0178, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0172, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.12526027858257294, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0137, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.15774008631706238, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0153, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.1049765795469284, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0155, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09307108074426651, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0186, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1369069367647171, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0158, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.12320058792829514, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0181, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.10885081440210342, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0168, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.08989177644252777, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0153, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.09219451993703842, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0172, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.11661440879106522, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0192, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.0971342995762825, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0178, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.10453730076551437, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0165, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.17652514576911926, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0196, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10733854025602341, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0154, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.0989762544631958, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0181, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.12107723951339722, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.1385297179222107, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0191, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.12209209054708481, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0217, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.09332361072301865, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0158, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.12448447197675705, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0164, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.12061425298452377, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0169, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.12132656574249268, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0183, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.14007116854190826, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0167, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.11448334902524948, + "learning_rate": 1.893057716200171e-05, + "loss": 0.015, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.12702956795692444, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0152, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.09236849099397659, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0177, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.09818986803293228, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.0164, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12131274491548538, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0171, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10609138011932373, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.0147, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.1431477814912796, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0173, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.0844314768910408, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0162, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.09800038486719131, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0157, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.11529140919446945, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0172, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.086520716547966, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0154, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.12068255245685577, + "learning_rate": 1.888667760339568e-05, + "loss": 0.015, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.12278824299573898, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0174, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.10242115706205368, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0173, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.11696609109640121, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0187, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.11133866757154465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0151, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.1319858729839325, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0165, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.08355733752250671, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0166, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.10663259029388428, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0177, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.09787473827600479, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0154, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.12111647427082062, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0171, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.10036861151456833, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.15351398289203644, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.019, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.11561891436576843, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0174, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.09197179228067398, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0174, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.0961032509803772, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.019, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.09264974296092987, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0158, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.11056753247976303, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0159, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.10464371740818024, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0165, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.13888190686702728, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0185, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.09801805019378662, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0189, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.15648017823696136, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0167, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1129915863275528, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0166, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.10456246137619019, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0171, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.10873697698116302, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.0168, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.17470459640026093, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0169, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.15504027903079987, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.10821744054555893, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0152, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.14031334221363068, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0181, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.128667414188385, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0183, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.09095419198274612, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0151, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.09303424507379532, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0167, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.10706666111946106, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0174, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.13011431694030762, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0177, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.13821059465408325, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0154, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.10680826753377914, + "learning_rate": 1.875076403503638e-05, + "loss": 0.017, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.12847846746444702, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0146, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.1472046673297882, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0184, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10622651875019073, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0153, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.129075825214386, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0189, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.12048091739416122, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.017, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.10954522341489792, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0175, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.10811172425746918, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0173, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.13378627598285675, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0193, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.13185591995716095, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0195, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.1013568863272667, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0182, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.07558644562959671, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0173, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.10668148100376129, + "learning_rate": 1.870271530813327e-05, + "loss": 0.018, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.134077250957489, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0157, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.14325739443302155, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0221, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.1389995515346527, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0178, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.13043907284736633, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0169, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.12082460522651672, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0174, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11521385610103607, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0165, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.11550573259592056, + "learning_rate": 1.867466815081593e-05, + "loss": 0.017, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.11315014213323593, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0173, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.1067388728260994, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0161, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.0941673144698143, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0158, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.09685055166482925, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0161, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.11744765937328339, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0164, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.10552993416786194, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0154, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.12654545903205872, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0184, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.11333387345075607, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0186, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.09856864809989929, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.0187, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.0933985561132431, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0171, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.096091628074646, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0194, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.08463368564844131, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0165, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.10648614913225174, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.017, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.10691184550523758, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0165, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1005169153213501, + "learning_rate": 1.861452118736741e-05, + "loss": 0.0174, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.1037023589015007, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0172, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.1299487054347992, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.015, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14639562368392944, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0167, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.12270855903625488, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0173, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.09706515073776245, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0169, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.11888926476240158, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.0152, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.11289773136377335, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0175, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13422279059886932, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0168, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.10992032289505005, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0176, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.10719195753335953, + "learning_rate": 1.85743888359202e-05, + "loss": 0.017, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.10348568856716156, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0166, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.10436776280403137, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0172, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.09779159724712372, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.018, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.10799746960401535, + "learning_rate": 1.855832827763672e-05, + "loss": 0.016, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.11972185224294662, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0195, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.10282087326049805, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0169, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.11840159446001053, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0183, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.13861460983753204, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0205, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10774466395378113, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0163, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.14446309208869934, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0168, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.1313662976026535, + "learning_rate": 1.853021190846946e-05, + "loss": 0.0161, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.0857798308134079, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0159, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.11933837085962296, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0178, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11762946099042892, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0154, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.11128528416156769, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0198, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.07360108196735382, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0148, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.10680225491523743, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0174, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.11387799680233002, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0188, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.2774786651134491, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0165, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.12314122170209885, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.017, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.14403507113456726, + "learning_rate": 1.849002290456207e-05, + "loss": 0.0175, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.10811382532119751, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0173, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.1399172991514206, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0154, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.10577643662691116, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0177, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.10794928669929504, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0155, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.0844842791557312, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0149, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.09744124114513397, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0166, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.14749610424041748, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0179, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.09801949560642242, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0177, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.07795082777738571, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0174, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.11197780072689056, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.0172, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.11819472163915634, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.08461402356624603, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0171, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.11102861911058426, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0156, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.11057187616825104, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0189, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.11031147837638855, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0155, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.08435497432947159, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0187, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.12450679391622543, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.017, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.09348208457231522, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0171, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.10584745556116104, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0161, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.14194349944591522, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.0177, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.1025288999080658, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0153, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.07800278812646866, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0157, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.1253945231437683, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0176, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.11724173277616501, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.016, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.13043718039989471, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0179, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.08728159964084625, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0177, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.1419975906610489, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0185, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.12774096429347992, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0187, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.10763564705848694, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.0163, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.12671130895614624, + "learning_rate": 1.836929808825424e-05, + "loss": 0.0177, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.10261370241641998, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0145, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12106029689311981, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.0157, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.12428469955921173, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0194, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.08797722309827805, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0175, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.14663904905319214, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.018, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.12288253754377365, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0176, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.11406450718641281, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0164, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10584931075572968, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.017, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1212797686457634, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0176, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.10806597024202347, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0159, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.10098059475421906, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0186, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.14030125737190247, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0204, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.09907040745019913, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0158, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1299905776977539, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0198, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.134843647480011, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0157, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.1048227846622467, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0154, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.12262892723083496, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0193, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.12806116044521332, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0167, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.12875764071941376, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.0149, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.09743807464838028, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0149, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10791302472352982, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0155, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10845673829317093, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0187, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.1366390436887741, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0161, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.101824089884758, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0174, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.12938271462917328, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0159, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.13503864407539368, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0196, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10241002589464188, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0153, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.10051950812339783, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.0159, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11316224932670593, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0166, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.10802372545003891, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0167, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.1365254521369934, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0176, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.11756277829408646, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.016, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13363559544086456, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0167, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.13138234615325928, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0194, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.1519714891910553, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0147, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.12152288109064102, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.017, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.0976492166519165, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0174, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.09356466680765152, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0158, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.09382101893424988, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0175, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.12188217788934708, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0163, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.09245157241821289, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0183, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.1661074161529541, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0202, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.154639333486557, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0143, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.117487333714962, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.017, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.09468945115804672, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0171, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.09632660448551178, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0147, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.12390028685331345, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0177, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.11924377828836441, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0154, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.13191300630569458, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0154, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.07403766363859177, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0171, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.10911701619625092, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0159, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.11888141930103302, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0189, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.09766717255115509, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0182, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.08285287767648697, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0145, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.13635197281837463, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0166, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.11627981066703796, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.10899969190359116, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0174, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.10002870857715607, + "learning_rate": 1.813525107780979e-05, + "loss": 0.016, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09963744133710861, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0163, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.2032729536294937, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0188, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.10394410789012909, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0178, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.1004093736410141, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0181, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.08776404708623886, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.0188, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.10926580429077148, + "learning_rate": 1.811099258182467e-05, + "loss": 0.018, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.09496597945690155, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0153, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.12430521845817566, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0152, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.11681065708398819, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0156, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.10450825840234756, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0162, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10932883620262146, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0163, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.1358717978000641, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0152, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.11744217574596405, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0188, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.14966122806072235, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0179, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.113619364798069, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.08801224827766418, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0168, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.09556800872087479, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0171, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.11116721481084824, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0171, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.10650353878736496, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0158, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.09222083538770676, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0154, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.10958139598369598, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.0189, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11020398885011673, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0166, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.08026755601167679, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0159, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.09629099816083908, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0162, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11310935020446777, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0192, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.1281706690788269, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0191, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12402708828449249, + "learning_rate": 1.802602130181036e-05, + "loss": 0.0159, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.09760920703411102, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0151, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.12342818826436996, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.018, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09423290938138962, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0163, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.10432913154363632, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0172, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.09628765285015106, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0169, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.11590366810560226, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0167, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10148236900568008, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0155, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.10599179565906525, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0155, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.11854606866836548, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.10626410692930222, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0185, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.12172172963619232, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0179, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.10842256247997284, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0173, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.11806859076023102, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0163, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11489225924015045, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0178, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.11859364807605743, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0164, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.18860413134098053, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.0168, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.13375963270664215, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0168, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.13163933157920837, + "learning_rate": 1.795310800633073e-05, + "loss": 0.0151, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.11135797202587128, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0175, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.12392686307430267, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0161, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.0995136946439743, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0165, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.10128366202116013, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0161, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.13080070912837982, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0169, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.10262267291545868, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0176, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.10983351618051529, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0163, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.09834928810596466, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0158, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.11170317232608795, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0166, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.14308896660804749, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.09110505878925323, + "learning_rate": 1.790851399590954e-05, + "loss": 0.017, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.12270113825798035, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0165, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.1422351896762848, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0177, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1294422596693039, + "learning_rate": 1.789634733729274e-05, + "loss": 0.017, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.1157895103096962, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0192, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.12473416328430176, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0156, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.1765025109052658, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0161, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11862069368362427, + "learning_rate": 1.78801220478228e-05, + "loss": 0.017, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.10549142211675644, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0165, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.09818614274263382, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0171, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.13607585430145264, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0168, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.09340739995241165, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0158, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.0805685818195343, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.0207, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.11172603815793991, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0164, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.08757263422012329, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0179, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.09590335935354233, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0199, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.1186867505311966, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.0146, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.09467568248510361, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0175, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11578786373138428, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0179, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09604159742593765, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0173, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.11337848007678986, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.0156, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.11909373849630356, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0188, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.08804072439670563, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0175, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.13454975187778473, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.0184, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.12201394885778427, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0148, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.11946159601211548, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0153, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12322071939706802, + "learning_rate": 1.78030044678245e-05, + "loss": 0.0181, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.10444212704896927, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0159, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.10723353922367096, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0158, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.10750993341207504, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0167, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.1207774356007576, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0207, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.12730205059051514, + "learning_rate": 1.778269750931628e-05, + "loss": 0.017, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.08926276862621307, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0151, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.10322589427232742, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0174, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10034044831991196, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.0184, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.12131599336862564, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0183, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.10715801268815994, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0156, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.14093570411205292, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0169, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.124595507979393, + "learning_rate": 1.775425890028875e-05, + "loss": 0.017, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.13153409957885742, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0189, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.1348494291305542, + "learning_rate": 1.77461316962853e-05, + "loss": 0.019, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.10926292091608047, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0161, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.11094044893980026, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0155, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.1341349333524704, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0184, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.15732111036777496, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0203, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.07597270607948303, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.0184, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.11097310483455658, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.0176, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.12087065726518631, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0164, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12998445332050323, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0191, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.13928557932376862, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0157, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.10487896203994751, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0172, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.13437892496585846, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0182, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.20842674374580383, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0189, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.09813268482685089, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12833304703235626, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0168, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.1019122451543808, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0159, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.07276969403028488, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.016, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09712188690900803, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0169, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.10403666645288467, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0156, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.11877460777759552, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0152, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.10731421411037445, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0188, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.12387207895517349, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0173, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.10561594367027283, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0179, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.2772648334503174, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0168, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.097810760140419, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0142, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.1168239563703537, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0171, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.14195922017097473, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0162, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11385945230722427, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0187, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.1124262586236, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.13074515759944916, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0151, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.12591321766376495, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0177, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.09103676676750183, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0169, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.12457307428121567, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0162, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.11226223409175873, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0156, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.14778819680213928, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0187, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.10433483868837357, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0187, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.08798553794622421, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0163, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.09729046374559402, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0149, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10188018530607224, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0154, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.12559989094734192, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.018, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12567821145057678, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0168, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.12211210280656815, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0166, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.11825426667928696, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0168, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.11676980555057526, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0175, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.11447525024414062, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0162, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.08958559483289719, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0161, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.11528483778238297, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0162, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.08296159654855728, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.014, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.12306056916713715, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0158, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.13315193355083466, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.0164, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.09533339738845825, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0166, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.20049314200878143, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.017, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.14355438947677612, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0187, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.11443490535020828, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0146, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.10336155444383621, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0164, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11926902830600739, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0173, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.09699229151010513, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0176, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.11406200379133224, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0164, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.10817506164312363, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0165, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.11398246139287949, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.0163, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.09427376091480255, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0166, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.10372194647789001, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0148, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.09003781527280807, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0172, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.09807098656892776, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.11895686388015747, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0163, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10839524865150452, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0173, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.20792271196842194, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0183, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.07746267318725586, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0147, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.11470094323158264, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0177, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.1377800703048706, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.1247936338186264, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.016, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.1405080407857895, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0164, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.1318410336971283, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0175, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.14254692196846008, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0159, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.12803269922733307, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0146, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11180103570222855, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0169, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.16513720154762268, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0162, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.13401633501052856, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.0182, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.13510771095752716, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0175, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.12391090393066406, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0174, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1019999086856842, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0157, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.11906205862760544, + "learning_rate": 1.7416306237948e-05, + "loss": 0.0179, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11846558749675751, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.0154, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.12292807549238205, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0163, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10195571929216385, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0186, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.1032365933060646, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0163, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1241045668721199, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0173, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.10956405848264694, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.0163, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.12094716727733612, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.018, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.1353612244129181, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0161, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.0935104638338089, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0164, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.10712502896785736, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.0177, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.12715180218219757, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0148, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.09601753205060959, + "learning_rate": 1.736733684019633e-05, + "loss": 0.0167, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.11189524084329605, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0174, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.11025495082139969, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.016, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.1203494742512703, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0192, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.12150877714157104, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0143, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.11957377195358276, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0146, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.16482087969779968, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0181, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.10647732019424438, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0178, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.13345423340797424, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0155, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.13268540799617767, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0178, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.12209953367710114, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0161, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.11951886862516403, + "learning_rate": 1.73224254699117e-05, + "loss": 0.0213, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.12044327706098557, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0195, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.10611505806446075, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0169, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.12347753345966339, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0145, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13930706679821014, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0185, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.12702012062072754, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0167, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.1193593367934227, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0169, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.16778719425201416, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0179, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11699119210243225, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0159, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11571445316076279, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.08481728285551071, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0159, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.10007977485656738, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0155, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.12851299345493317, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0167, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.13929468393325806, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0154, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.11237664520740509, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0143, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.11654951423406601, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0179, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.11065302044153214, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0168, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.10024461150169373, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0173, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.1334248036146164, + "learning_rate": 1.724888829234145e-05, + "loss": 0.016, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.1115458682179451, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0159, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.15415695309638977, + "learning_rate": 1.72407140459686e-05, + "loss": 0.0166, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10965333878993988, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0159, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.10152038931846619, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0152, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.10030021518468857, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0161, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.1526176780462265, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0134, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13154515624046326, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.017, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.12260743975639343, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0152, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.07395336031913757, + "learning_rate": 1.721209883968656e-05, + "loss": 0.0155, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.1107548177242279, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0178, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.08634423464536667, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0148, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.10544566810131073, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0151, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.13896141946315765, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0171, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10351476818323135, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0164, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.11912661790847778, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0163, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.12101101130247116, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0185, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.09519081562757492, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0137, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.13252407312393188, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0153, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.08964477479457855, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0148, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.11772924661636353, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0153, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1261531263589859, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.0155, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.12295454740524292, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0162, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.14834828674793243, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0162, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.14458344876766205, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0146, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.10006255656480789, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0183, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11527495086193085, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0174, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.11176037788391113, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0146, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.11118462681770325, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0139, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.1216764822602272, + "learning_rate": 1.713029611235338e-05, + "loss": 0.018, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10572952032089233, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0155, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.09299932420253754, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0193, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.11654805392026901, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0146, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.125601127743721, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0148, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.12143341451883316, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0144, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.10944563895463943, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0188, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.10531438142061234, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.017, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.08527344465255737, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0134, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.12489438056945801, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0184, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.12359192967414856, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0162, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.09504237025976181, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0147, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.1144440770149231, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0155, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.11942129582166672, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.0161, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.15496809780597687, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0167, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.11641886830329895, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0155, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.13092041015625, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0164, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.1662251204252243, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0169, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.1151951253414154, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0146, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.13008327782154083, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.017, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.08816810697317123, + "learning_rate": 1.70484286724112e-05, + "loss": 0.0147, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.12602686882019043, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0167, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10360333323478699, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0161, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.10288142412900925, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0174, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.1306314915418625, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0178, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.11859587579965591, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0152, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.12666994333267212, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.0155, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.10859639197587967, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0174, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.1423543244600296, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0145, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.12444380670785904, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0165, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.07491649687290192, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0174, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.12173314392566681, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0156, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.11790800094604492, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.1375807374715805, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0182, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.11894135922193527, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0149, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.09377862513065338, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0157, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1448003053665161, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0189, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.14296171069145203, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.019, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.10862060636281967, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0188, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.1081303209066391, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.0166, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08262617141008377, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0163, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12013579159975052, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0161, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.06014449894428253, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0166, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09688957780599594, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0164, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.11730696260929108, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.0142, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.11243025213479996, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0134, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11974845081567764, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0156, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.1408282220363617, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0193, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.1495579332113266, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.017, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.1408172994852066, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.19624975323677063, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0165, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09940195083618164, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0168, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.11277889460325241, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0171, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.13027498126029968, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0165, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.11570801585912704, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.0159, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.12621626257896423, + "learning_rate": 1.690501241902545e-05, + "loss": 0.016, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.13131381571292877, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0146, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.11577112972736359, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0168, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.1240323930978775, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0162, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.17400938272476196, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.0155, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.09539356827735901, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0165, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10108646005392075, + "learning_rate": 1.688040861067922e-05, + "loss": 0.016, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.1295158863067627, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0151, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.13638637959957123, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0188, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.12255847454071045, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0148, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.13242852687835693, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0166, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11082842200994492, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.015, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.14538930356502533, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0172, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1609591394662857, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0166, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.10611626505851746, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.0142, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.13291358947753906, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0179, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.1112443208694458, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0179, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.10251478105783463, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0174, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.10443990677595139, + "learning_rate": 1.683118563837068e-05, + "loss": 0.016, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.08651736378669739, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0176, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.09996092319488525, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.017, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.13269121944904327, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.0154, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.121453657746315, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0167, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.10587169975042343, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0161, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.12990498542785645, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0158, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.12159401923418045, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0158, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.12046501785516739, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0165, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.10626068711280823, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0144, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.14602169394493103, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0161, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09537681937217712, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0188, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.1014745756983757, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0171, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.11358807235956192, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0165, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.12290225178003311, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0153, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.11178604513406754, + "learning_rate": 1.676962882515376e-05, + "loss": 0.0175, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.1464385688304901, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0155, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.12778854370117188, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0144, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.09556689858436584, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0171, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.10866247862577438, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.0151, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.14120467007160187, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0163, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.12688933312892914, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0174, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.10251957923173904, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0171, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.11465741693973541, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.016, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.08181405812501907, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0201, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.09379345923662186, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0172, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.12122894823551178, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0183, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.0956249088048935, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0142, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.11953603476285934, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0169, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.12494538724422455, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0167, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.20808497071266174, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0138, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.10776039958000183, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0152, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.09606330841779709, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0154, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.10988017171621323, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0182, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.10328707098960876, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0146, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.166867196559906, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0179, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.11701484769582748, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0159, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.10510280728340149, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0179, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.12857040762901306, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0172, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1627577394247055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0176, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.10962127149105072, + "learning_rate": 1.666696746039845e-05, + "loss": 0.0147, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.12961529195308685, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0164, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.08055070787668228, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0148, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.10041354596614838, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0161, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.12196766585111618, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0138, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.0879533439874649, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0156, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.1102190688252449, + "learning_rate": 1.664231677578745e-05, + "loss": 0.016, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.12463485449552536, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.016, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.10829959809780121, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0181, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.08602527529001236, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0176, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.10241527855396271, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.0172, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11007986962795258, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0156, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13110534846782684, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.018, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1417531818151474, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0161, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.13233700394630432, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0147, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.09148544073104858, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0145, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.08781567215919495, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.0149, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.10577751696109772, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0161, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.11689155548810959, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0171, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.0777219608426094, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0152, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.07983718812465668, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0158, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1595761626958847, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0168, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.1037725880742073, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0161, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.10239970684051514, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0149, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.15451422333717346, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0199, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.11882807314395905, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0165, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.09850180149078369, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0147, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.10899021476507187, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.09254994243383408, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0148, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.13559140264987946, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0162, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.11541012674570084, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0154, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.07937203347682953, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0153, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.10062690824270248, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0165, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.13341091573238373, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0178, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.0857294499874115, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0132, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.10551585257053375, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0142, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.12363193184137344, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0165, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.11550596356391907, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0165, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.09379533678293228, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0154, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.10079154372215271, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0163, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.13103018701076508, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0178, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.1172979474067688, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0152, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.11038312315940857, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0152, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.0925338938832283, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0148, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.2519476115703583, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.017, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.09703299403190613, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0193, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.10749004036188126, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0175, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.08736859261989594, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0147, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.13125599920749664, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0161, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10677823424339294, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0156, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.09135954827070236, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0151, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.10548236966133118, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0179, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.1150064468383789, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0188, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.127374529838562, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.016, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.09152884781360626, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0177, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.1021062359213829, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0145, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.1054648831486702, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0172, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.12128820270299911, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0165, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.11375363171100616, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0151, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.11004973948001862, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0182, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.11101444810628891, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0164, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.10393026471138, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0172, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.1139075756072998, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0159, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1569976508617401, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0178, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09649650752544403, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0165, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.16198694705963135, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0157, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.10348767787218094, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0173, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.12259576469659805, + "learning_rate": 1.639145830609088e-05, + "loss": 0.0171, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.10480271279811859, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0159, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.11254257708787918, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0163, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.11534930765628815, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0149, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.10350872576236725, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0169, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.10812029242515564, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.015, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.11796360462903976, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.0144, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.1276526153087616, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0157, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.11665219813585281, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0151, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.10940699279308319, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0165, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.09596745669841766, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0149, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15527229011058807, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0164, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11651212722063065, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0145, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.12004360556602478, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0153, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.13298730552196503, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0176, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.1496417224407196, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0153, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.11104541271924973, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0158, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.12215209007263184, + "learning_rate": 1.632147383062092e-05, + "loss": 0.015, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.09335075318813324, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0148, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.11481857299804688, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0153, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14369069039821625, + "learning_rate": 1.630912057457699e-05, + "loss": 0.019, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.11009083688259125, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0162, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.11112431436777115, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.0137, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.10135089606046677, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0167, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09966760873794556, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0168, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12062226980924606, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0145, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.1109282448887825, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0162, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.10862662643194199, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0152, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.09910363703966141, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0156, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.1374131590127945, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0167, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.12114953994750977, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.0156, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.14391179382801056, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0156, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.12364984303712845, + "learning_rate": 1.625969868712804e-05, + "loss": 0.0169, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.08771048486232758, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0142, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.10959397256374359, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0176, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.138074591755867, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0149, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.1116161122918129, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0154, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.1398090422153473, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0164, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.12145344167947769, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0173, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.11162104457616806, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0153, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.1211397722363472, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.016, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09488996863365173, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0146, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.11799386143684387, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0137, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.1406150758266449, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.0149, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.1307811290025711, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0192, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.13287024199962616, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.0153, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10396360605955124, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0133, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.11891327053308487, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0167, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.12922705709934235, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0178, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10563277453184128, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0132, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.17530488967895508, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0159, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.10669656842947006, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0166, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.09745346754789352, + "learning_rate": 1.617729852865125e-05, + "loss": 0.0152, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.12519428133964539, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0136, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.10747594386339188, + "learning_rate": 1.616905650452066e-05, + "loss": 0.0147, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.10352121293544769, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0161, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.1046888679265976, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0131, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.2108439952135086, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0148, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.1401670128107071, + "learning_rate": 1.61525713933778e-05, + "loss": 0.0171, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.10765372961759567, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0169, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.14557960629463196, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0151, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.11173100024461746, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0181, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.08689062297344208, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0156, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.10052058845758438, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0171, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.11383800208568573, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0167, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10390642285346985, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0157, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.10242070257663727, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0147, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.10699810087680817, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.0162, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.1349048614501953, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0167, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.10787097364664078, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.0144, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.06869964301586151, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0168, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.08449438959360123, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.0156, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10576694458723068, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0157, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.10042312741279602, + "learning_rate": 1.609073991709496e-05, + "loss": 0.018, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.10223086178302765, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0179, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.11301153153181076, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0151, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.11257045716047287, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0159, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.09129524976015091, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0154, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.10601308941841125, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0157, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.11378258466720581, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0151, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.1138806939125061, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0188, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.10367786139249802, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0157, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.13393838703632355, + "learning_rate": 1.605363201758077e-05, + "loss": 0.016, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.09532077610492706, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0145, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.13368268311023712, + "learning_rate": 1.60453849294712e-05, + "loss": 0.0178, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10675307363271713, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0148, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.10722523182630539, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0167, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.09825004637241364, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0148, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.14284753799438477, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.0194, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0167, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.13374702632427216, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0153, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.08964364975690842, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.0164, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.07354595512151718, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0138, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.08927512913942337, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0167, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.10888224840164185, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0193, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.12273052334785461, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0157, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.10807470232248306, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0152, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.1124584972858429, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.0148, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.10052672028541565, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0145, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.08666819334030151, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0172, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.11364969611167908, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0155, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.11263672262430191, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0164, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.1379120796918869, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0187, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.10857756435871124, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0133, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.13408084213733673, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0156, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.10847405344247818, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0141, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.10151072591543198, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0138, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11481117457151413, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0138, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.10202853381633759, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0153, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.09370319545269012, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.0159, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.09684473276138306, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0162, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.11064363270998001, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0162, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.15720140933990479, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0167, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.12132807075977325, + "learning_rate": 1.592576773310894e-05, + "loss": 0.0159, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.10888595134019852, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0133, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.0892426148056984, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0163, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12278196215629578, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0186, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10831111669540405, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0155, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.08722463995218277, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0156, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.11901626735925674, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0155, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.08866825699806213, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.0155, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.09303046017885208, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0152, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.12453419715166092, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0166, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.11046329140663147, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0155, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.10482461750507355, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0154, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11235640943050385, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0165, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.11749114841222763, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0159, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.1201302632689476, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0174, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.13044941425323486, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0143, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.10600301623344421, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0155, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.12556208670139313, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0159, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.12454571574926376, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0153, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.13055634498596191, + "learning_rate": 1.584736531020125e-05, + "loss": 0.0188, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.1379219889640808, + "learning_rate": 1.584323820456826e-05, + "loss": 0.016, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.13937926292419434, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0181, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.21694405376911163, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.013, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.11387006938457489, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0144, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.10102641582489014, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0154, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.10343687981367111, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0161, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.14018288254737854, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0175, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10792441666126251, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0168, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.10692388564348221, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0175, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.11248136311769485, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0176, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.13729289174079895, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0154, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.1389625370502472, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0152, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11402085423469543, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0154, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.14007720351219177, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0151, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.10529543459415436, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0151, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.09746415913105011, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0136, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1150866374373436, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0158, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.10207786411046982, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0139, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.13951019942760468, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0157, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11709082126617432, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.0158, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.12401671707630157, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0145, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.103446826338768, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0149, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.16494140028953552, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0192, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.10893384367227554, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0141, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.09813039004802704, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.0142, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.14741374552249908, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0184, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.12035968154668808, + "learning_rate": 1.573591189751e-05, + "loss": 0.0157, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.0970459058880806, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0176, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.10946980863809586, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0144, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.15628819167613983, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0157, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.12485401332378387, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0184, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.11361467838287354, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0163, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.19872187077999115, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0152, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.0912356749176979, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0167, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.09560416638851166, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0178, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.10187338292598724, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0127, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.11934805661439896, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0145, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.11719757318496704, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0148, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09480439126491547, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0167, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.09031171351671219, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0163, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.12236128002405167, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.014, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.09773887693881989, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.015, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.11999719589948654, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.0145, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.11341046541929245, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0142, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.10415472090244293, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0183, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.10177286714315414, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0152, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.10778342932462692, + "learning_rate": 1.565332720308093e-05, + "loss": 0.016, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.08542678505182266, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0157, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.10727313905954361, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0164, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.1413055658340454, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0163, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.10777363926172256, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0175, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.14286576211452484, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0155, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.10413800925016403, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0164, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.10655509680509567, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0163, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.10989648103713989, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0153, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.09240605682134628, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0144, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.0842762365937233, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0143, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.13544097542762756, + "learning_rate": 1.560789700123175e-05, + "loss": 0.017, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11756332218647003, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0159, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.11726696044206619, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0165, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.11693235486745834, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0146, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.1371903419494629, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0145, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.0971926674246788, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.017, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.08220933377742767, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0156, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.10194069892168045, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0134, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.08653450012207031, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.0131, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.08990923315286636, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0144, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.10370788723230362, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0133, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.06916087865829468, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.0139, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.1136801689863205, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0146, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12992094457149506, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.0144, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.09605134278535843, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0165, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.10273651778697968, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0165, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.11571857333183289, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0151, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.10102164000272751, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0146, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.11045269668102264, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.0146, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.12626338005065918, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0166, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.1357768476009369, + "learning_rate": 1.552528261458322e-05, + "loss": 0.018, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.10587792843580246, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0143, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.1509760171175003, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0187, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11034028977155685, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0174, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.07888679206371307, + "learning_rate": 1.550875774213821e-05, + "loss": 0.0159, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.11219298094511032, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.13006654381752014, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.0161, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12485881894826889, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0148, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.10320480912923813, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0171, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.12367358803749084, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.019, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1412898451089859, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0196, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.10843265801668167, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.0146, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09489759802818298, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0147, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.10236489772796631, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0158, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.12652312219142914, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0173, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10095653682947159, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0167, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.0802559182047844, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0145, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.1239897757768631, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0155, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.09682597219944, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.0154, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.10482888668775558, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.013, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.1272023320198059, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0164, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.09988776594400406, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0166, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.10352584719657898, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0142, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.36612969636917114, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0169, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.12097200751304626, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09437882900238037, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.0147, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14130210876464844, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0147, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.1284940391778946, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0148, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12240798771381378, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0135, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.08258320391178131, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0154, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.10075270384550095, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0162, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.10424498468637466, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0138, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.10447008907794952, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0156, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.10173123329877853, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0127, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.12066177278757095, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0162, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.09489274024963379, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0159, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.09174267202615738, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0155, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.13935479521751404, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0158, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.11126208305358887, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0142, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.13917426764965057, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.014, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.10231155157089233, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0141, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.11477960646152496, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0145, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.10886958241462708, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0155, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.09005019068717957, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0147, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.09915502369403839, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0138, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.1025547906756401, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0158, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.1030225157737732, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0158, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.0842776820063591, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0163, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.0634579285979271, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0151, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.09493366628885269, + "learning_rate": 1.532281505788039e-05, + "loss": 0.014, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12343619018793106, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0167, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.10681836307048798, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0149, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.1302853226661682, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0171, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.10506479442119598, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0136, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.12051218003034592, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0168, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11805758625268936, + "learning_rate": 1.52980183248759e-05, + "loss": 0.0158, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.10309842228889465, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.015, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.108099564909935, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0159, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11023534834384918, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0136, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.11571311205625534, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0163, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.1068359762430191, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.017, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.15622149407863617, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0156, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.11074791103601456, + "learning_rate": 1.526908777801797e-05, + "loss": 0.0163, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.11449336260557175, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0156, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.1219671219587326, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.0167, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.08449103683233261, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0162, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.13202930986881256, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0167, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.11665860563516617, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0173, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.13584332168102264, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0148, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.09913904964923859, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0155, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09704537689685822, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0162, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.13063789904117584, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0153, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.1202530562877655, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0182, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.08398830890655518, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.015, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.09059721976518631, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0129, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.10968997329473495, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0141, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.09898033738136292, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.1568402498960495, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0161, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.09885324537754059, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0168, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09452448785305023, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0142, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.16158591210842133, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0149, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.13292880356311798, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.016, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.17052333056926727, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0165, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.11351195722818375, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0163, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.09296253323554993, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0128, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.09396881610155106, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0157, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.13142332434654236, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0161, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.1246478259563446, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0135, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.13347554206848145, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0173, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.10838479548692703, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.11035843938589096, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0158, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.10406306385993958, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0149, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.16070416569709778, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0149, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09154598414897919, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.014, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.10438812524080276, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0164, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.09018818289041519, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0149, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1320919543504715, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0151, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.0870271846652031, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0155, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.12103381007909775, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0191, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.09007984399795532, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.0141, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.10315215587615967, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.19641031324863434, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0169, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14303788542747498, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0207, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.10933230072259903, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0136, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.11214675009250641, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0175, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.09008116275072098, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0145, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.09065063297748566, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.016, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.1116153672337532, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.016, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.09231071174144745, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0148, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12521710991859436, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0159, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.12240038067102432, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0143, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.09896460920572281, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0154, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.10471562296152115, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0152, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.11139113456010818, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0132, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.09953293949365616, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.015, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.12128512561321259, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0155, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.14064094424247742, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0132, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.09164810925722122, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0144, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.11127574741840363, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0145, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.11218927800655365, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0168, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.10342293232679367, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0137, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.10499373823404312, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0151, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.08029376715421677, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0157, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09445271641016006, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0161, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.09650483727455139, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0155, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.08300688117742538, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0176, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.11890654265880585, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0145, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10451909154653549, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0195, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.08795508742332458, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0141, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.1659373790025711, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0197, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.1001206561923027, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0161, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10190606862306595, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0161, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.11775147914886475, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0139, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.11382468789815903, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0149, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10219351947307587, + "learning_rate": 1.497147766283292e-05, + "loss": 0.0143, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.0930316224694252, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0148, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.10227128118276596, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0153, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.10634683817625046, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0163, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.1443759799003601, + "learning_rate": 1.495494301522528e-05, + "loss": 0.016, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.10867374390363693, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0147, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.11999388039112091, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0157, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10372292995452881, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0161, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.14905284345149994, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.0172, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10412498563528061, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0145, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.09797366708517075, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0133, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.1329793930053711, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0179, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.09716201573610306, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0146, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.09868425875902176, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.0153, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.0929884985089302, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0159, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.1417207568883896, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0151, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.13194335997104645, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0177, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.11850415170192719, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0173, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08122166991233826, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0149, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.11195199191570282, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.12405557930469513, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0135, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.12344473600387573, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0153, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.11776174604892731, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.0154, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.11634577810764313, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0158, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.11508087068796158, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0165, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.12009026855230331, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0141, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.15844415128231049, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.014, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.11803163588047028, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.014, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.11727108061313629, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.0157, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.09923852235078812, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0153, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.14936032891273499, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0174, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.10241764038801193, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0164, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.11330186575651169, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.016, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.14860020577907562, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0152, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.1636815220117569, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.015, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.11093862354755402, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0142, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.11333172768354416, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0141, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.10221018642187119, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0155, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.11085604131221771, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0158, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.17324623465538025, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0153, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.10631918907165527, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.0136, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.09826772660017014, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0148, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.1288551241159439, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11162615567445755, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0148, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.1307016909122467, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.017, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11215298622846603, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0154, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.09471828490495682, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0145, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10456835478544235, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.0147, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.08179501444101334, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0129, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.11810965836048126, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0151, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.0977158397436142, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0139, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.09650871902704239, + "learning_rate": 1.476067039948277e-05, + "loss": 0.0138, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.10262744128704071, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0145, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11392423510551453, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0147, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11900563538074493, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0163, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.10666237771511078, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0162, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.10767115652561188, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0152, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.1232513040304184, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0144, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.12271980196237564, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0167, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.12620051205158234, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0176, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11799243837594986, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0153, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.11633779853582382, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0154, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.10304611176252365, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0173, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.11186012625694275, + "learning_rate": 1.471107401829042e-05, + "loss": 0.0143, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.12054259330034256, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0174, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.1436678171157837, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0146, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.12709002196788788, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0171, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.11617328971624374, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0146, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.1117766946554184, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0178, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.12944459915161133, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0155, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11261707544326782, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0144, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.11154191195964813, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.013, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.11878378689289093, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0142, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.12532871961593628, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0176, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.10492552816867828, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0165, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.10776105523109436, + "learning_rate": 1.466148079674631e-05, + "loss": 0.016, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.10712877660989761, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0165, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.091713547706604, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0155, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.10679589211940765, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0185, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.11124882102012634, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0141, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.10930055379867554, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0148, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.10555030405521393, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0152, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.0836755633354187, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0153, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.11849679052829742, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0158, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.13186359405517578, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.0167, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.11089541018009186, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0189, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.09473541378974915, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0155, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.1151607483625412, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0138, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.10416971892118454, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0138, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.10269354283809662, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0158, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11249537020921707, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0152, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.14352446794509888, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0169, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10745783895254135, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0159, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.09522922337055206, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0164, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.13701501488685608, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0146, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.08020906895399094, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0139, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09878123551607132, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.015, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.09399155527353287, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.015, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.11354538053274155, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0175, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.09135859459638596, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0156, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.08270548284053802, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.014, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.11252828687429428, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0157, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.0985439196228981, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0163, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.12125496566295624, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0146, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.09888023883104324, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0176, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.10250411182641983, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.10701503604650497, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0151, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11525820940732956, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0157, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.09755205363035202, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0159, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09472718834877014, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.015, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10360235720872879, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.0149, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.09946251660585403, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0164, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.12811636924743652, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0135, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.127360999584198, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.016, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.12480203062295914, + "learning_rate": 1.45003312024933e-05, + "loss": 0.0142, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.10674259066581726, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0156, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.12259232252836227, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0149, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.08998828381299973, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0149, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.10316051542758942, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0145, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.14545974135398865, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0158, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.1868322491645813, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.015, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.12345045059919357, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0154, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.11794525384902954, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0177, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.12257082760334015, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0152, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.12155821174383163, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0159, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12235453724861145, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0147, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.1056676059961319, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.015, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.1309097409248352, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0145, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.11743862181901932, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0152, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.12476198375225067, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.016, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.08397012203931808, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0162, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12207506597042084, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0159, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.09591932594776154, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0137, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.12573577463626862, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0164, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1315455585718155, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0164, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.1468929648399353, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0165, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.15008211135864258, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0148, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.09235277771949768, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0153, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.1275569051504135, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0163, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.08151543140411377, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0129, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10390543937683105, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.015, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.11043227463960648, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0141, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.1271016001701355, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0163, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09329107403755188, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0134, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.10737643390893936, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0135, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12903659045696259, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0159, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.11276338994503021, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0147, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.14026683568954468, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0163, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.12503503262996674, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0172, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.10071003437042236, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0138, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10627497732639313, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0138, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.0770413801074028, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0135, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.09400150179862976, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0154, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.1549134999513626, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.015, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.09969423711299896, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.0135, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.09770094603300095, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0149, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.0848713293671608, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0128, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08489260822534561, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0142, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.14606203138828278, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.0144, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.1283080279827118, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0147, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.09238846600055695, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0154, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.09167003631591797, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0153, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.1081148013472557, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0142, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.19600129127502441, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0158, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.08589469641447067, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0148, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10786846280097961, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0123, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.12004593759775162, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0154, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10496815294027328, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0133, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.1324186772108078, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0152, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.10160262882709503, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0137, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.14738032221794128, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0167, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.1030116155743599, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0146, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.10793277621269226, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0156, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.11900396645069122, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0165, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.12115667760372162, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0144, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.12846702337265015, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0163, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.09553872793912888, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.015, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.12807151675224304, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0161, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.12759897112846375, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.017, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.10901748389005661, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0151, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.09706093370914459, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0129, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10346078872680664, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.0147, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.12170220911502838, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0136, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.24494236707687378, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0139, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.10071292519569397, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0131, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.13647668063640594, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0151, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09296777099370956, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.0147, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.1015978679060936, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.0147, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.09014145284891129, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0134, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.08798721432685852, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0136, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.08899430930614471, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.014, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.10661277174949646, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.0189, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.10300924628973007, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.0165, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.10107792913913727, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0152, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.12540587782859802, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0147, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.14788442850112915, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0154, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.1628517359495163, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0155, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.12338890135288239, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0156, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09656467288732529, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0132, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.15762092173099518, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0142, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.1834212690591812, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0156, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.091162770986557, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.0142, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.14656929671764374, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0163, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.08447254449129105, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0144, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10864629596471786, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0137, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.12742409110069275, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0134, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.09852556139230728, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.13655197620391846, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0143, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.09299396723508835, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0134, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.08303989470005035, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0136, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.09860356897115707, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.016, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.08682689070701599, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0134, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.09411480277776718, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.0149, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.10258188843727112, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.016, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.09782259166240692, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0165, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.10300199687480927, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0148, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.09208223968744278, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0175, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.09144087135791779, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0148, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.09090441465377808, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0153, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.10571702569723129, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0168, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.1070566177368164, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0154, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.12103649973869324, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.015, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.10678276419639587, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0165, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11873649060726166, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0154, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.11791808903217316, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0145, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.0986858382821083, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.016, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09249810129404068, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0127, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12782645225524902, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0154, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.10502453148365021, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0162, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.09005366265773773, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.0143, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11655479669570923, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0158, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.10148292779922485, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0155, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.11018893122673035, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0147, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.11230522394180298, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0148, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.09911413490772247, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0159, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.09358150511980057, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0154, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.12689724564552307, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0145, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.10737589001655579, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.013, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.12670603394508362, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0148, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.1459159404039383, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0162, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.14686602354049683, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0139, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10709210485219955, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0161, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.08311928808689117, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.015, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12497760355472565, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0161, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.08203094452619553, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0129, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.12578807771205902, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0145, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10697296261787415, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0147, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.10980327427387238, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0154, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.09966560453176498, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0126, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.09740720689296722, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0126, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.1452278196811676, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.013, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.10068526864051819, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0152, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.10273153334856033, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0134, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.1101747378706932, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0148, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.11141166090965271, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0133, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1535777896642685, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0155, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.08465548604726791, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0161, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.07614064961671829, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0149, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09852079302072525, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0135, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.13796597719192505, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0163, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1669035404920578, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0143, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09881193935871124, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0142, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.12325254082679749, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0132, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.16859179735183716, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0168, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.09475059062242508, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.014, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.11981360614299774, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0186, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.10775262862443924, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0151, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.10053553432226181, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0129, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.10618098825216293, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0154, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.12225139886140823, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0171, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.07435912638902664, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0152, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.10211736708879471, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0177, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12564173340797424, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.015, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.10412495583295822, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0157, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.12016817182302475, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0146, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.10957430303096771, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0141, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08035340160131454, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0146, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.1221131756901741, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0153, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.10435835272073746, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0133, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.11107036471366882, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0137, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.0575222373008728, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0149, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.076436348259449, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0159, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.09325376152992249, + "learning_rate": 1.380704261284659e-05, + "loss": 0.0163, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.11715361475944519, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0161, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.08567722141742706, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0149, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.12162303924560547, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.017, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.11919409781694412, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.0154, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.10694293677806854, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0138, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.10736183077096939, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0133, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11603212356567383, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0127, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.1180153489112854, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.0161, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.11968789249658585, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.0151, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.1250530481338501, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0136, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10276933014392853, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0174, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.14728401601314545, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.0138, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.09268351644277573, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0129, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12262973934412003, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0147, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.11068148910999298, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0157, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.09095525741577148, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0149, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.12430994212627411, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0154, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1246076300740242, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0153, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.08052714169025421, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.0144, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.1667853593826294, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0164, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.10500273108482361, + "learning_rate": 1.372053087451907e-05, + "loss": 0.013, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.10832768678665161, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0171, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.11245214939117432, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0158, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.14754413068294525, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0146, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.16368411481380463, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0175, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.1533215045928955, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0138, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.11878912895917892, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0134, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.10748414695262909, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0155, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.14132313430309296, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0139, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.09151305258274078, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0133, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10014014691114426, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0143, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.12393905967473984, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0154, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13815879821777344, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0155, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.11488667875528336, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.014, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.08504864573478699, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0149, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.12521564960479736, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0161, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.09525968134403229, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0137, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.12103450298309326, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.017, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.10642504692077637, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0139, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.15175847709178925, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0153, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11594653874635696, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.017, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.12788720428943634, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0171, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.12372581660747528, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0157, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.09206387400627136, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0155, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.09841199964284897, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0162, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.135525643825531, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0148, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.0982549637556076, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.0148, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.07432328909635544, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0124, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09830821305513382, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0139, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.10874123871326447, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0149, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.08979356288909912, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0189, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.08131087571382523, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.0128, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.1204582154750824, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0141, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11884505301713943, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0144, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.16451388597488403, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0134, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.11953870207071304, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0135, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12111926078796387, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0159, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.1181047335267067, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0142, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.12142453342676163, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0156, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09890097379684448, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.0137, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.13294243812561035, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0132, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.1397709846496582, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0161, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.10867296159267426, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.014, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11157155781984329, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.015, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09516428411006927, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0156, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.21431738138198853, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0158, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.09854292124509811, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0159, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.21810708940029144, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0152, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.14076393842697144, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0162, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.0892275720834732, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0154, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.09736334532499313, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0127, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.13880307972431183, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.015, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1091906875371933, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0147, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11605603992938995, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0168, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.11140985041856766, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0162, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.09057680517435074, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0161, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.12626419961452484, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0161, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.136195570230484, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0142, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.12415066361427307, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0121, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.12581513822078705, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.0151, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.11835133284330368, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0147, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.12255530059337616, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.0135, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10576335340738297, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0149, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.12575241923332214, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.014, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09812266379594803, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0152, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.11845924705266953, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0143, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.10975389182567596, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0147, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.09649760276079178, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0131, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.12912778556346893, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.0152, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10225717723369598, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0143, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.1712195724248886, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0144, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.1477300226688385, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0155, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.11692417412996292, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0146, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.1232081949710846, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0152, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.14507624506950378, + "learning_rate": 1.341604094324139e-05, + "loss": 0.017, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.13909533619880676, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0136, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.1438206136226654, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.0167, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.09637414664030075, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0134, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.15815401077270508, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.015, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.12961018085479736, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0159, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.11171260476112366, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0133, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.0907273143529892, + "learning_rate": 1.33872699666556e-05, + "loss": 0.015, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.11181363463401794, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0162, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.11502614617347717, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0169, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.14428892731666565, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0158, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.1440858244895935, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0134, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.07517627626657486, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0129, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.11884543299674988, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0147, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.08753574639558792, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0178, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.09831796586513519, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.014, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.12262003868818283, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.016, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.11444123834371567, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0142, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11559469252824783, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0138, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.09936029464006424, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0116, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.12810789048671722, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.0137, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.07923603802919388, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0136, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.09709552675485611, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0148, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.0924353376030922, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0152, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.10444460809230804, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0136, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.09510108828544617, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.014, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.11014958471059799, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.014, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.13198788464069366, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0153, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.09857133030891418, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0128, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.15041643381118774, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0148, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.13857626914978027, + "learning_rate": 1.329277958632556e-05, + "loss": 0.018, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.13004948198795319, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0142, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.09864714741706848, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0143, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12682363390922546, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0177, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.09303848445415497, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0141, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.13559986650943756, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0153, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.10143767297267914, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0131, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.1287422627210617, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0142, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.12000516802072525, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.016, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.10434254258871078, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.014, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08141680061817169, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0138, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08781418204307556, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0138, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.1355503350496292, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.015, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.1243978887796402, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0143, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.11088231950998306, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0154, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12559033930301666, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0132, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.16149592399597168, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.09809687733650208, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.0142, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.12586510181427002, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0165, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.12403467297554016, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.14458157122135162, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.0135, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.09313284605741501, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0165, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.14337806403636932, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0144, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.12650802731513977, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0126, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.112495556473732, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.0154, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11324562877416611, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.0139, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.11720716208219528, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.0146, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.1261121928691864, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0143, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.1302812099456787, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0157, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.12074442952871323, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0172, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.11716746538877487, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0158, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.11565797030925751, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0144, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.11653099209070206, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0155, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.13103044033050537, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0152, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.10341408103704453, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0138, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.11098334193229675, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.0146, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.11139992624521255, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.10786942392587662, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0143, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.10139934718608856, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0154, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.12619639933109283, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0153, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.13168810307979584, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0154, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.09545969218015671, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0151, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.09984135627746582, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0129, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10903751850128174, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0135, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.11882342398166656, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0151, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.1105988621711731, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0137, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.09874245524406433, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0142, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.10673530399799347, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0157, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.10264459252357483, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0162, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10769633203744888, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0133, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.16478870809078217, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0149, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.12630558013916016, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0128, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09824837744235992, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0144, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12136032432317734, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0124, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.10943865031003952, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0154, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.11394404619932175, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.0117, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.10907739400863647, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0144, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10671219974756241, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0156, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.10080806165933609, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0141, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.17395640909671783, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0145, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.12620997428894043, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0186, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.13434381783008575, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0149, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.09921885281801224, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.0149, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.13742485642433167, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0155, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.0998004674911499, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10389069467782974, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0138, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.11996421217918396, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.0147, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.09136862307786942, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0129, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.11111150681972504, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0159, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.11024610698223114, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0138, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.15376223623752594, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0154, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.13934020698070526, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0138, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.11208872497081757, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0131, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1513150930404663, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.10351566225290298, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0137, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11078765988349915, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0164, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.15622863173484802, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.0147, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.11603175848722458, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0145, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.1254235804080963, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0162, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.0900387316942215, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.015, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09852428734302521, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0172, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.13502047955989838, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.014, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.1460445076227188, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0156, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.10317493230104446, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0129, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.127616748213768, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0157, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.11560963094234467, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0142, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.11239839345216751, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0135, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.1503314971923828, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0157, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.1253356635570526, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0154, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.13217659294605255, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0127, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09087327867746353, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0144, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.09290582686662674, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0147, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.10273045301437378, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0165, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.14741986989974976, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.016, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.0804770216345787, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0125, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.12413053959608078, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0136, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.09822124987840652, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0176, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.09091541171073914, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0161, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.09487786144018173, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0148, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1009742021560669, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0147, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.08731451630592346, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0118, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.08123810589313507, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0134, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.10559827089309692, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0159, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.11313227564096451, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0164, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.11703694611787796, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0171, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.1083052009344101, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0168, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.13762544095516205, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0185, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.07979343086481094, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.09789150208234787, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0139, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.1905684620141983, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0158, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.1036764457821846, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0118, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.09413912892341614, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0137, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.2435447722673416, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.0132, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.09290791302919388, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0146, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.1211535856127739, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0151, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.1133081242442131, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.013, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.11362706869840622, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0171, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.10019481182098389, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0151, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.09489522129297256, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0143, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12912064790725708, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0138, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.11348122358322144, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0136, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.1325748860836029, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0156, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.11978204548358917, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0146, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.10022053122520447, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0137, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.07388417422771454, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0152, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.10844552516937256, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.0142, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.12381294369697571, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0132, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09625408798456192, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.0154, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.1207597553730011, + "learning_rate": 1.276827841111951e-05, + "loss": 0.016, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.12216818332672119, + "learning_rate": 1.276419082873553e-05, + "loss": 0.016, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.0969374030828476, + "learning_rate": 1.27601034161465e-05, + "loss": 0.0125, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.13043035566806793, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0145, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.10600738227367401, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0144, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.11113949120044708, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0135, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.09458228200674057, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0139, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10076881945133209, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0139, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1297597885131836, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0142, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.10059160739183426, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0142, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10716533660888672, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0149, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.1138077974319458, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0151, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.13693730533123016, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0136, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.08157949149608612, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0143, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.09702442586421967, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0121, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09029403328895569, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0133, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.1444508135318756, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0133, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.11026642471551895, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0161, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.11229082942008972, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0135, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.08556455373764038, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0166, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10684676468372345, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.0142, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.11970192193984985, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0132, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.10227229446172714, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.015, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.08301195502281189, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0153, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.1275268793106079, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.014, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.11613821983337402, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.015, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.10491939634084702, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0127, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.09623268246650696, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0136, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.111306332051754, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0137, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0863005742430687, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.0137, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.09434706717729568, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0142, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.10771634429693222, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0155, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.14157304167747498, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.0145, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11419932544231415, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0154, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.12308389693498611, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0133, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.10775505006313324, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.0144, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.10383253544569016, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0141, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.1012682244181633, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0144, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.12071908265352249, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0157, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.12535089254379272, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0149, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.08634085953235626, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0152, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.10035993158817291, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0138, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.13245077431201935, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0135, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.08713636547327042, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0144, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.10049428045749664, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0134, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.10432936251163483, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0163, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.16578327119350433, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0166, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.16106748580932617, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0125, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.11968962103128433, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0142, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1189485490322113, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0155, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.10795646905899048, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0139, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.08580315858125687, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0134, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.17969708144664764, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0159, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09675301611423492, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0146, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.10921840369701385, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0158, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.11278913170099258, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0133, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12176436185836792, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0138, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.1186685562133789, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0171, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.10296562314033508, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.015, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.15394653379917145, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0173, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.06815146654844284, + "learning_rate": 1.252333462101563e-05, + "loss": 0.014, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.1100328117609024, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0129, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.08578243851661682, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0145, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.11262256652116776, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.015, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.09644307941198349, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0183, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09589987248182297, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0145, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.09658264368772507, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0144, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.11963947862386703, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0118, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.14156748354434967, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0164, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.14387111365795135, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0151, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.11542942374944687, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0123, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.12162452191114426, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0165, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.13585466146469116, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0157, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.11272095143795013, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0159, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.1828201562166214, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.0162, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.09920861572027206, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0163, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.12618015706539154, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.016, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.13111864030361176, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0155, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.08486807346343994, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0126, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.14975349605083466, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0156, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.12495499104261398, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0154, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.11145244538784027, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0147, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.08242745697498322, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0149, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.10702498257160187, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0152, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.16176027059555054, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0145, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.0959012433886528, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0134, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10282868891954422, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0143, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.11747556924819946, + "learning_rate": 1.241332681912234e-05, + "loss": 0.0165, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10770756006240845, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0134, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.09925016760826111, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0126, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.09207499027252197, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.016, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.21546097099781036, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0131, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.10848663747310638, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0182, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.09382427483797073, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0134, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.10758832842111588, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0136, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.10075095295906067, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0154, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.09366318583488464, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0151, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.13190370798110962, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0154, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.1926490068435669, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0162, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.10018754750490189, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0127, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10250869393348694, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0146, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.08466067910194397, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0152, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.10397373139858246, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.0148, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14372816681861877, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0151, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11059669405221939, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0133, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09743087738752365, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0151, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.11265920102596283, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0145, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.08748266100883484, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0146, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.10545652359724045, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0133, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.09680107235908508, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0149, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.13201938569545746, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0152, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.12063595652580261, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0167, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.09986719489097595, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0153, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.09913598746061325, + "learning_rate": 1.230752865632506e-05, + "loss": 0.0151, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.1373804211616516, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0144, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.12088140845298767, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0137, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.08200784027576447, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0145, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.12470699846744537, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0154, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.10114862769842148, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.13169868290424347, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0159, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.08321736752986908, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0147, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.10108216106891632, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.012, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.08475375175476074, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0155, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.08946104347705841, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0145, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.07863342016935349, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0143, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.0906776711344719, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0123, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.11063490062952042, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0128, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.10001818835735321, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0133, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.12135739624500275, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0145, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.1233866810798645, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0133, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.09280609339475632, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0161, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.11677120625972748, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0144, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.11756281554698944, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.0163, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.08084359019994736, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.1415908932685852, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0149, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09107301384210587, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.013, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.11712655425071716, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.014, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.08366917073726654, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0125, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10391885787248611, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0146, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.1042654886841774, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0132, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.09504523873329163, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0148, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.12793844938278198, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0164, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.16210904717445374, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.15506179630756378, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.0153, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.11274553835391998, + "learning_rate": 1.218156580082751e-05, + "loss": 0.015, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.14852653443813324, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0172, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13558489084243774, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0156, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.08902081102132797, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0152, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12315832078456879, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0154, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.13528762757778168, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0145, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.1538013368844986, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0151, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.09894628077745438, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0154, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07696802169084549, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0136, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.09229687601327896, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0131, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.11709824949502945, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0146, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09195996820926666, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0142, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.11714771389961243, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.0157, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.11561911553144455, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0162, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.12904389202594757, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0129, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.10347782075405121, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0126, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.10054155439138412, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0126, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.1713176816701889, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0142, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10279975831508636, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0133, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.11513496935367584, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0137, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.096926249563694, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0123, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.1159687489271164, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.0145, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09745503962039948, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.014, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.15608133375644684, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.0144, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.08617804199457169, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.09285406023263931, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0123, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.11964505165815353, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0143, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1376206874847412, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0136, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.12900781631469727, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0148, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.10527875274419785, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0157, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.11285959929227829, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0135, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.09756818413734436, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0158, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.10695865005254745, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09360404312610626, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0129, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.100432388484478, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0126, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.10982918739318848, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0158, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.07767278701066971, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.015, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.10380839556455612, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0153, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.1225915253162384, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0167, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.0981304869055748, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0139, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.10055239498615265, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0146, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.09311894327402115, + "learning_rate": 1.201528720560666e-05, + "loss": 0.016, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.11859868466854095, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0163, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.15336444973945618, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0153, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.09175246208906174, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0125, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.0973501205444336, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0155, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.1643284410238266, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0142, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.1077687218785286, + "learning_rate": 1.19909852066355e-05, + "loss": 0.015, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.07722865790128708, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0125, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.149483323097229, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.015, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.11405838280916214, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0144, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.11132190376520157, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0142, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.10609457641839981, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0128, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.09458910673856735, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0134, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.10288971662521362, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0164, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.0970534011721611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0128, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.09571509063243866, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0129, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.08922260254621506, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0144, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0886709988117218, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.017, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.11732205003499985, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0141, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.11602790653705597, + "learning_rate": 1.193835918571248e-05, + "loss": 0.016, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.11643379926681519, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.0156, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.0927688404917717, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0149, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11533524841070175, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0133, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11001960933208466, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0135, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.12629783153533936, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.014, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.13336558640003204, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0121, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.11672402173280716, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.016, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.10014890134334564, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0136, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.0940401554107666, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.0151, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.1272585242986679, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0139, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.08492123335599899, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0151, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.1195707842707634, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0137, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.1095622330904007, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0167, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.09656022489070892, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0137, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1358073204755783, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0166, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.10981698334217072, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0141, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.07301078736782074, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0127, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.1040695309638977, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0143, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.16375352442264557, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.0136, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.19000792503356934, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0172, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11925465613603592, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0148, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.188803032040596, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.0151, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.09543585777282715, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0129, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11177625507116318, + "learning_rate": 1.184130718987356e-05, + "loss": 0.013, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.11703382432460785, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0153, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.12326977401971817, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0151, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.09018218517303467, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0147, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1641266793012619, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0146, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.08780542016029358, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0127, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09138940274715424, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0133, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.11768870800733566, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0128, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10298353433609009, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0141, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.10553838312625885, + "learning_rate": 1.180494808593063e-05, + "loss": 0.014, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10878341645002365, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0166, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10571706295013428, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0166, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.09349539875984192, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0126, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.10237022489309311, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0128, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.11938492953777313, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0167, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.10417849570512772, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0132, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.10322093963623047, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0141, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.09445950388908386, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.0146, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.1363389790058136, + "learning_rate": 1.176860863606944e-05, + "loss": 0.014, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.10975046455860138, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.013, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.12059775739908218, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0136, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.10529737174510956, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0158, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.11280884593725204, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0146, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.09665497392416, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0126, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.14059090614318848, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0142, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.10380115360021591, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.0132, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.10641331225633621, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0157, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.10845568776130676, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0135, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.10574200004339218, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0153, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.09128011763095856, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0142, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.10704760998487473, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0129, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.1180003359913826, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0141, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.11784005165100098, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0138, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.08203493803739548, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0144, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.10125461220741272, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.0142, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.12113944441080093, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0124, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.09266798943281174, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0156, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.11875846236944199, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0146, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.11041444540023804, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.013, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.13168254494667053, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0151, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.08208412677049637, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0132, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.08591622114181519, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0152, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.14605054259300232, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.014, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.12816093862056732, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0163, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.10751980543136597, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.013, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.1012972742319107, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0134, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13583813607692719, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0121, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.13573803007602692, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0143, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11719135195016861, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.0163, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.11966057121753693, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0133, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.11172359436750412, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0146, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10357552021741867, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0134, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.09571186453104019, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0143, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.13586629927158356, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0162, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.10461284965276718, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0134, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.182363823056221, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.014, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.12570780515670776, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0161, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.10319641977548599, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0136, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1066908910870552, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0123, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.10409434139728546, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0143, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.08230862766504288, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0128, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.110751211643219, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0148, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.09656667709350586, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0137, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.10626243054866791, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0127, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10294152796268463, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0151, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.09571471065282822, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0136, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.1162625253200531, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0129, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.09755981713533401, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0151, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.08586591482162476, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.016, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.11462593078613281, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.013, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.11438485980033875, + "learning_rate": 1.155904354214324e-05, + "loss": 0.016, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.15895701944828033, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0161, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.1135995015501976, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0128, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.11993709206581116, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0142, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.14105743169784546, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0118, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.08521834015846252, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0136, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.17094957828521729, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0159, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.1379796713590622, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.012, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.10417291522026062, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0138, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.09216596931219101, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0151, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.08398481458425522, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0124, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.1291581243276596, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0152, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.09349840134382248, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0127, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.08282706141471863, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0143, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.10378136485815048, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0146, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1504564881324768, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0144, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.0940636470913887, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0147, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.11142407357692719, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0127, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.16776388883590698, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0158, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.10488522052764893, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0132, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.09612040966749191, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0119, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.09318249672651291, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0116, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.08647014945745468, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0124, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.1067274957895279, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0136, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.08162420988082886, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0136, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.08949539810419083, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0136, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.11327249556779861, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0158, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.10963857173919678, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0141, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.10124490410089493, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.014, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10530558228492737, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0135, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.1251840591430664, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0127, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.14418582618236542, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0156, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.09544084221124649, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0137, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.09794268012046814, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0151, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.10741079598665237, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0127, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.10632490366697311, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0117, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08640208840370178, + "learning_rate": 1.141437135633602e-05, + "loss": 0.0119, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.09476958960294724, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0134, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.1141064390540123, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0147, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.1008499413728714, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0119, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.0853564441204071, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.12970876693725586, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0128, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11144055426120758, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0137, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.09443596750497818, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0127, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.19357536733150482, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0173, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.11861733347177505, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0145, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.13152548670768738, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0145, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11607249826192856, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0128, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.09456845372915268, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0145, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.12110379338264465, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.014, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.07872474938631058, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0128, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.1042177751660347, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0136, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.15168064832687378, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0126, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.09023087471723557, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0125, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.1110723614692688, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0121, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.09889500588178635, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0136, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.13798406720161438, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0151, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.14507459104061127, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0175, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10034530609846115, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0143, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.09978580474853516, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0142, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.11770957708358765, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0145, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.10433546453714371, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0148, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.1006259173154831, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.016, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.09567620605230331, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0133, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08685444295406342, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0116, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.10901349037885666, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.0137, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.15114818513393402, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0147, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.08353722840547562, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0146, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.14387859404087067, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0135, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.12125595659017563, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0148, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.11398079991340637, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0142, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.14072677493095398, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0149, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.11925073713064194, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0153, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.10301142185926437, + "learning_rate": 1.126604838049735e-05, + "loss": 0.013, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.11580533534288406, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.012, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.10776577144861221, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0143, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.1724127233028412, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0146, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.15091492235660553, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0128, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.10073890537023544, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0131, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11136317253112793, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0127, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.11012828350067139, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0133, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.12291619181632996, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0148, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.15791606903076172, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0144, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.12378063052892685, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.0137, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12597671151161194, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0172, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.1783512681722641, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0169, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.07375630736351013, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0139, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.07327539473772049, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0125, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.09731314331293106, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0144, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.0931868702173233, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0124, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.10977564752101898, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0132, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.0881611704826355, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0168, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.1239505335688591, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0133, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.10325739532709122, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0132, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11499917507171631, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0119, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.10131840407848358, + "learning_rate": 1.117803938925701e-05, + "loss": 0.014, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.10435229539871216, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0119, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.10805722326040268, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0136, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0759270042181015, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0167, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.1406354457139969, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0155, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1452491581439972, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.10772743821144104, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0148, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09850231558084488, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0155, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.08501581847667694, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0155, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11223053187131882, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0146, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.07870364189147949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0136, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.11768847703933716, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.0153, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.096342071890831, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0137, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.10153007507324219, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0139, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.0837850570678711, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0117, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.11286687850952148, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0148, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.14671675860881805, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0117, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.09201597422361374, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0161, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.11397483944892883, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0127, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.08533219248056412, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0126, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.18251192569732666, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0134, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.17834120988845825, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.014, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11550542712211609, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0127, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11998283863067627, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0149, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11872667074203491, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0155, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.10344526916742325, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.015, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.11288250237703323, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0149, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.16417740285396576, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0128, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09344452619552612, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0131, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11792241781949997, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0159, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10204951465129852, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0143, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.13064894080162048, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0151, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.08712288737297058, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0148, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.11878253519535065, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0123, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0907600000500679, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0135, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.08273347467184067, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0143, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.08222810924053192, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0112, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.11497339606285095, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0166, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.115402452647686, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0123, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.09611446410417557, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0128, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.12478714436292648, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0133, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.1713663637638092, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0118, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.13887059688568115, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0144, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.08571378141641617, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0154, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.1155383437871933, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0172, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.1123170331120491, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0153, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.13517709076404572, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0142, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.10638624429702759, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0121, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.11200621724128723, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0129, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10821404308080673, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.0128, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.10840310156345367, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0119, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.10352504998445511, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0144, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.10084252059459686, + "learning_rate": 1.097058241668254e-05, + "loss": 0.0128, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.11881101131439209, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.09931256622076035, + "learning_rate": 1.09626195543497e-05, + "loss": 0.0148, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.1260150521993637, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0125, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14202995598316193, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0133, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.15601785480976105, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0158, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.10701733827590942, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0139, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10180657356977463, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0138, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.0962422713637352, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.0124, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.13017651438713074, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0153, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.1295875608921051, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0154, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.10589175671339035, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0144, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.10173171758651733, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0142, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11211177706718445, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.0144, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.09875764697790146, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.014, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.11233554035425186, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0135, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.11327439546585083, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0136, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.10059859603643417, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0143, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.1160704717040062, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0132, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.15698356926441193, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0163, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.08235745877027512, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0134, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.12417344003915787, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0145, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.25982916355133057, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0155, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.12248262017965317, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0139, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.11425210535526276, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.015, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.12354899197816849, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0187, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.1102118268609047, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0137, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11539250612258911, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.015, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.12405017763376236, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0136, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.10405998677015305, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.014, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.0918804481625557, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0162, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.10429271310567856, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0128, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.09917982667684555, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0133, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.15390615165233612, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0133, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.11222631484270096, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0151, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.14449205994606018, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0165, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1332763433456421, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.015, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09963493794202805, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0132, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.12338289618492126, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0122, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.09369654208421707, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0127, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.13490226864814758, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0149, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.08246663957834244, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0158, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.0892438217997551, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0124, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.1275915652513504, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0135, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.150055930018425, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0118, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.08317377418279648, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0141, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.118314728140831, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.0131, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.086687833070755, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0128, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.10310143977403641, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0125, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.06559336930513382, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0135, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.11680027842521667, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0135, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.1381155252456665, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.015, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.08212172240018845, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0139, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.11066312342882156, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0145, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.11565965414047241, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0131, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.08814681321382523, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0122, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.11116618663072586, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0156, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.10491067916154861, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0157, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.15800951421260834, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0141, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.10169819742441177, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0156, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.07719215005636215, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0127, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.08958051353693008, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0145, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.1171223521232605, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0141, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.09662648290395737, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0121, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.09235699474811554, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0129, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.127174973487854, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0157, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.10298880189657211, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0146, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08706901967525482, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0132, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.10898775607347488, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0117, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.12178633362054825, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0129, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.10881572961807251, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0138, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.12326113879680634, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.013, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.13331058621406555, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0129, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.11921956390142441, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0154, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.10736659914255142, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.014, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09894116222858429, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0141, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.09651704877614975, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0146, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.11836998164653778, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.0176, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.1108580157160759, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0129, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.1281266212463379, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0138, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.13581599295139313, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0149, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10911770164966583, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0162, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.14131440222263336, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0126, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.10660075396299362, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0144, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.12547564506530762, + "learning_rate": 1.063722589416218e-05, + "loss": 0.0142, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11465640366077423, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0143, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.1274002492427826, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0145, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.10782597959041595, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0144, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.08273816108703613, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0126, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.12750636041164398, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0113, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.10962793231010437, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0132, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11728726327419281, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0146, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.10677216202020645, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0146, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.0974075049161911, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.0129, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.09190504997968674, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0129, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.10864423215389252, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0122, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.08833710849285126, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0141, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.14306439459323883, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.013, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.11269631236791611, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0115, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.08850698918104172, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0116, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10025947540998459, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0116, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.07444360852241516, + "learning_rate": 1.057003955600053e-05, + "loss": 0.013, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.10606181621551514, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0133, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.11394638568162918, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0143, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.12015067040920258, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0128, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.11465302109718323, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0117, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.09164674580097198, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0116, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.10532126575708389, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0128, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.16767282783985138, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0137, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.16740484535694122, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0151, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08568812906742096, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0135, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.24568453431129456, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0138, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.13317807018756866, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0142, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11310465633869171, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0139, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.09677092730998993, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0123, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.13912886381149292, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0152, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.0902378037571907, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0124, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09430436044931412, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0116, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.13361187279224396, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0161, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09588316082954407, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0135, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.13104380667209625, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0161, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.11476746946573257, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0137, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.087586909532547, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0124, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.12136878073215485, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0156, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.09384764730930328, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0137, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.10422758758068085, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0142, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.11391496658325195, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0125, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.12965570390224457, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0137, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.13042569160461426, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.0159, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.10433127731084824, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0129, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.0989370346069336, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0132, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.08399394899606705, + "learning_rate": 1.045171352558859e-05, + "loss": 0.013, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.11693604290485382, + "learning_rate": 1.044777463361479e-05, + "loss": 0.0148, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.12304933369159698, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0121, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.10788783431053162, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0138, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11043704301118851, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0135, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.11343314498662949, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0158, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.10075241327285767, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0143, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.08752909302711487, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.012, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.11036614328622818, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0126, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.09241441637277603, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.0157, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.1369611620903015, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.0132, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.12782007455825806, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0136, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.140136256814003, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0144, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11508572846651077, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.0136, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.14334864914417267, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0138, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.10154326260089874, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0135, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.11290924996137619, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0134, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.13472969830036163, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0144, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.1215982660651207, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0125, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.10628669708967209, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0125, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.07950089126825333, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0135, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.14542941749095917, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0145, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.1189931258559227, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0115, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.11141669005155563, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0135, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.08776746690273285, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0136, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10942544043064117, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0139, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.15493649244308472, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0123, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.1485723853111267, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.0141, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.0886952131986618, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0124, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.09021062403917313, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0132, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.17778685688972473, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0126, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.11892827600240707, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0125, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.14066243171691895, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0135, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.10870271921157837, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0147, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.10636188089847565, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0156, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.11178954690694809, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.014, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.09059949964284897, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.0127, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.11947090923786163, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0145, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.1133369505405426, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.014, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.12451290339231491, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0139, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10219574719667435, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0138, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.09843198210000992, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0144, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.09784907102584839, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0155, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1259097307920456, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0156, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.08604080229997635, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0113, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.11517608165740967, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0131, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.09721313416957855, + "learning_rate": 1.027088684228894e-05, + "loss": 0.011, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.08404546976089478, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0164, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.11207510530948639, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0151, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.1101691946387291, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0129, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.2047414630651474, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0143, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.08887176960706711, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0139, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.08608511090278625, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0162, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10840177536010742, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0133, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.1228431761264801, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.0173, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.09171874076128006, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0119, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.1357734203338623, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0151, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.13374143838882446, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0134, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09826765209436417, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0112, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.30271241068840027, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0128, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.1474897861480713, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.0142, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.11385788768529892, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0121, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.1145087331533432, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0125, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.15277381241321564, + "learning_rate": 1.020425036775911e-05, + "loss": 0.011, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10591624677181244, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0123, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.09938671439886093, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0136, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.12457235157489777, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0145, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.10550948232412338, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0127, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.12241430580615997, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.013, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.12996822595596313, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0127, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.10886852443218231, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0124, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.10862760245800018, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0124, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.12071339040994644, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0152, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.12002567946910858, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0144, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.0731884092092514, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.014, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.09087152034044266, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0122, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.14147141575813293, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0126, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.10024842619895935, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.013, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.12605364620685577, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0129, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.07398458570241928, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.012, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.09070646017789841, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0139, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.06930165737867355, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.012, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.10134197026491165, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0131, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.12521374225616455, + "learning_rate": 1.012598938117799e-05, + "loss": 0.012, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.10238701850175858, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0139, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.10689418017864227, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0138, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.0954018235206604, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0125, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.1394306868314743, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0125, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.11217380315065384, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0143, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.1513020396232605, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.0122, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.11155983060598373, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0132, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.1257336288690567, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.015, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.10818599164485931, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0163, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.15947701036930084, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0142, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.10929335653781891, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0129, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.08994319289922714, + "learning_rate": 1.007910370568197e-05, + "loss": 0.011, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.09995368868112564, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0131, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.10167150944471359, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.0142, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1480625420808792, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0129, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.11566069722175598, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.013, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.10494371503591537, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0132, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.09531164914369583, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0116, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.0887216180562973, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0133, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.09848874062299728, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0122, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.14705902338027954, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0135, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.1412736475467682, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11742296814918518, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0143, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.11000296473503113, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0144, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.111237533390522, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0125, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.1085008904337883, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0133, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.12815521657466888, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0138, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.11128585785627365, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.12439922988414764, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0139, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.11863227933645248, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0154, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.13517265021800995, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0128, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.0960967019200325, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0123, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.0953422263264656, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0121, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.1965440809726715, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.0924379900097847, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0112, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.08345641940832138, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0146, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.10248678177595139, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0121, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.2980184257030487, + "learning_rate": 9.977703371038337e-06, + "loss": 0.013, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.11192528903484344, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0133, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.21329663693904877, + "learning_rate": 9.969913958481688e-06, + "loss": 0.0116, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.0922258049249649, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0159, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.12964527308940887, + "learning_rate": 9.962126073932371e-06, + "loss": 0.0143, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.11362354457378387, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0137, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.0902123972773552, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0126, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.08770131319761276, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0123, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.09604862332344055, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0131, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.08182749897241592, + "learning_rate": 9.942663063115248e-06, + "loss": 0.013, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.17937488853931427, + "learning_rate": 9.938771611983594e-06, + "loss": 0.0158, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.1343632936477661, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0139, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.14316903054714203, + "learning_rate": 9.93098986311647e-06, + "loss": 0.012, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11366818100214005, + "learning_rate": 9.927099565971976e-06, + "loss": 0.0129, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0949450209736824, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0144, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.13775621354579926, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0134, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.10221140086650848, + "learning_rate": 9.915430987239967e-06, + "loss": 0.012, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.11101876199245453, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0121, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.12785188853740692, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0146, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.1400655061006546, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0115, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.09468253701925278, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0122, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.08724251389503479, + "learning_rate": 9.895991082746168e-06, + "loss": 0.0143, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.10251675546169281, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.11980220675468445, + "learning_rate": 9.888217832192542e-06, + "loss": 0.015, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.12473925948143005, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0144, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.08260339498519897, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0122, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.11015481501817703, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0154, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1120007336139679, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0143, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09239251911640167, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0146, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10413528233766556, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0144, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.0863887220621109, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0143, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.0963623896241188, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0129, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.2090010792016983, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0155, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12460873275995255, + "learning_rate": 9.84937491900087e-06, + "loss": 0.012, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.11049462109804153, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0134, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.15180432796478271, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0142, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.13392038643360138, + "learning_rate": 9.837729659166156e-06, + "loss": 0.013, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.11534144729375839, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0134, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.1077096238732338, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0135, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.10903865098953247, + "learning_rate": 9.826087927694543e-06, + "loss": 0.014, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.09442178905010223, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0132, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.12171642482280731, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0143, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.08623645454645157, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0135, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.10249772667884827, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0151, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.16136185824871063, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0113, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.15169787406921387, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0161, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.095759317278862, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0143, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.11120729893445969, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0127, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09804334491491318, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0123, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.11462689191102982, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0111, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10223866254091263, + "learning_rate": 9.783431894635035e-06, + "loss": 0.012, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.1001039668917656, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0149, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.13274405896663666, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0136, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.09141503274440765, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0135, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.1291927844285965, + "learning_rate": 9.767932474193758e-06, + "loss": 0.0136, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10977526009082794, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0135, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.1167440190911293, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0138, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.08445189893245697, + "learning_rate": 9.756312079913248e-06, + "loss": 0.0129, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.2519867718219757, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0146, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.08974999934434891, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0152, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.1252872496843338, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0147, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.16116158664226532, + "learning_rate": 9.740823797373594e-06, + "loss": 0.013, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.08403292298316956, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0119, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.11238662898540497, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0154, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10106278210878372, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0136, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.08145973831415176, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0131, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.15683840215206146, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0132, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.14468176662921906, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0146, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.08105593919754028, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0122, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.12362894415855408, + "learning_rate": 9.709866422291219e-06, + "loss": 0.013, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.09933894127607346, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0156, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.09758161753416061, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0145, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.1307433545589447, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0139, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.1336813122034073, + "learning_rate": 9.694397367364635e-06, + "loss": 0.015, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.08841264992952347, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0162, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09142819046974182, + "learning_rate": 9.68666525627968e-06, + "loss": 0.0132, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11312894523143768, + "learning_rate": 9.682799805859474e-06, + "loss": 0.0129, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.14984415471553802, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0147, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.104182668030262, + "learning_rate": 9.675070116731382e-06, + "loss": 0.013, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.10078790783882141, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0131, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.08538635075092316, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0135, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.11059115827083588, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0134, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.10521579533815384, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0131, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.1130848377943039, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0137, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.09023619443178177, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0125, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.11787812411785126, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0136, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.14386765658855438, + "learning_rate": 9.6441675594234e-06, + "loss": 0.0153, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.10471580922603607, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0128, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.1340251863002777, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0139, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.1131536215543747, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0128, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10629499703645706, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0147, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.10359197854995728, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0121, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.11530483514070511, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0136, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.09336108714342117, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0124, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.08179444819688797, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0124, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.12341088801622391, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0125, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.12123409658670425, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0149, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.11406739801168442, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0141, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.07922708988189697, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0114, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.09156256914138794, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0121, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.09705155342817307, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0134, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.10157317668199539, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0157, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.09271441400051117, + "learning_rate": 9.582440688952473e-06, + "loss": 0.0113, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.09375326335430145, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0109, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.07997307926416397, + "learning_rate": 9.57473221124502e-06, + "loss": 0.0137, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.11743918806314468, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0138, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.09150344878435135, + "learning_rate": 9.567025381590669e-06, + "loss": 0.013, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.09968695789575577, + "learning_rate": 9.563172585515032e-06, + "loss": 0.011, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.10516709089279175, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0114, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.09716377407312393, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0132, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.12646295130252838, + "learning_rate": 9.551616675805319e-06, + "loss": 0.0145, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.11036434769630432, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0153, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10279295593500137, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0138, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.08990927040576935, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0132, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.10534624755382538, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0138, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.09507470577955246, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0114, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.101425401866436, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0137, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.11853020638227463, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0121, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.13193756341934204, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0128, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.10581472516059875, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0123, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.12424535304307938, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0139, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.1036083921790123, + "learning_rate": 9.509276927101462e-06, + "loss": 0.0126, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.07990803569555283, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0112, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.098060242831707, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0136, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.11892619729042053, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0126, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.10609585791826248, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0127, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.10905738174915314, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0123, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.0981525182723999, + "learning_rate": 9.486203760067013e-06, + "loss": 0.012, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.16955061256885529, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0119, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.10360468924045563, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0136, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.08537023514509201, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0112, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.1317787766456604, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0128, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10730036348104477, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0151, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09596534818410873, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0117, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.10839905589818954, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0138, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.12511657178401947, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0138, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.1078735962510109, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0127, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.10344097018241882, + "learning_rate": 9.447782011716307e-06, + "loss": 0.0119, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.18708902597427368, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0139, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.09599670022726059, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0125, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.13859155774116516, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0143, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09208878129720688, + "learning_rate": 9.432425102360072e-06, + "loss": 0.0126, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.21019718050956726, + "learning_rate": 9.428586931343613e-06, + "loss": 0.0117, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.09587983787059784, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0122, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.09626271575689316, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0124, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.1051478385925293, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0154, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.10166551172733307, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0119, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.09312118589878082, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0145, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.11722584068775177, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0126, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.0939759686589241, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0124, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.11434069275856018, + "learning_rate": 9.39789681972375e-06, + "loss": 0.012, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12582115828990936, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0137, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.07123387604951859, + "learning_rate": 9.390228540420876e-06, + "loss": 0.0155, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.08926647901535034, + "learning_rate": 9.386395039661539e-06, + "loss": 0.014, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.07371679693460464, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0128, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.1008603498339653, + "learning_rate": 9.378729317382715e-06, + "loss": 0.012, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.10988970100879669, + "learning_rate": 9.37489709644538e-06, + "loss": 0.0136, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.09886831045150757, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0123, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.10959886759519577, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0135, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.12443319708108902, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0161, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10207810997962952, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0146, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.11660879850387573, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0127, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.13432076573371887, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0142, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.11405819654464722, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0135, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.12952250242233276, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0119, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.09084811806678772, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0173, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.12252718955278397, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0099, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1293795108795166, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0144, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11287739127874374, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0164, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.1253224015235901, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0133, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.10046776384115219, + "learning_rate": 9.321290990574668e-06, + "loss": 0.014, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.08746109157800674, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0128, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.08306485414505005, + "learning_rate": 9.313639857014607e-06, + "loss": 0.0115, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.1340319961309433, + "learning_rate": 9.3098149378516e-06, + "loss": 0.0137, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.09587698429822922, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0143, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.14915677905082703, + "learning_rate": 9.302166396212059e-06, + "loss": 0.0143, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.09313366562128067, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0115, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.12605343759059906, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0146, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.12807077169418335, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0113, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.1186203733086586, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0119, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.10223432630300522, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0141, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.12712709605693817, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0132, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.12993593513965607, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0162, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.12273230403661728, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0133, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.1391134113073349, + "learning_rate": 9.26776941165236e-06, + "loss": 0.014, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.11695246398448944, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0137, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.1214752346277237, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0122, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.13961699604988098, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0117, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.1402655988931656, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0159, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.10683967918157578, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0146, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.08510883897542953, + "learning_rate": 9.244857665179213e-06, + "loss": 0.013, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.10308418422937393, + "learning_rate": 9.241040568803432e-06, + "loss": 0.0141, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.09736710786819458, + "learning_rate": 9.23722390978261e-06, + "loss": 0.0139, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.13270588219165802, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0118, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.08694430440664291, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0132, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.07487893849611282, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0123, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.1315941959619522, + "learning_rate": 9.221961653045449e-06, + "loss": 0.0115, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10982084274291992, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0146, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.09258668124675751, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.11609145998954773, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0143, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.1349419355392456, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0139, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.08282078802585602, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0108, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.12237592041492462, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0149, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.07864762842655182, + "learning_rate": 9.195269608854445e-06, + "loss": 0.012, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.1032051146030426, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0112, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.1346377581357956, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0132, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.08922939747571945, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0118, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.09849593788385391, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0141, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.07525688409805298, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0126, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.1287003457546234, + "learning_rate": 9.172407902473605e-06, + "loss": 0.0117, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.11016935855150223, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0123, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.09082796424627304, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0119, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.12462439388036728, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0133, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.10698465257883072, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0126, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.07600131630897522, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0133, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.09345563501119614, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0144, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.1186055988073349, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0151, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.12372780591249466, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0133, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.116783007979393, + "learning_rate": 9.138145235699377e-06, + "loss": 0.0139, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.08260510116815567, + "learning_rate": 9.134340494716546e-06, + "loss": 0.012, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.1815570890903473, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0125, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.12819825112819672, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0133, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.09238126873970032, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0115, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.09430833160877228, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0119, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.7733426094055176, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0128, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.10260160267353058, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0126, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.11146093159914017, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0136, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.11368793249130249, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0141, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.08438809961080551, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0134, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.11850261688232422, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.13420401513576508, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0128, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.09115619212388992, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0127, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.09974579513072968, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0141, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.10242673754692078, + "learning_rate": 9.081121025468816e-06, + "loss": 0.0133, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.08850608766078949, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0113, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.08285540342330933, + "learning_rate": 9.073525417675752e-06, + "loss": 0.0122, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.08347760140895844, + "learning_rate": 9.069728288749965e-06, + "loss": 0.011, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.12695780396461487, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0134, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.10132371634244919, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0143, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.07370374351739883, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0112, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.0993623211979866, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0138, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.07118473947048187, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0134, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.10804349929094315, + "learning_rate": 9.046954982950699e-06, + "loss": 0.014, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.12209299206733704, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0127, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.0896674394607544, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0134, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.10866492241621017, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0145, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.06327113509178162, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0125, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.1044771745800972, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0138, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10678671300411224, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0119, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.10146090388298035, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0122, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.11703896522521973, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0126, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.08116868138313293, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0117, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.16623066365718842, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0135, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.07943270355463028, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0128, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.17520980536937714, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0141, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.10311958938837051, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0124, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09508595615625381, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0139, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.11690590530633926, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0139, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.11585275083780289, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0119, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.08755052834749222, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0152, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.15034662187099457, + "learning_rate": 8.978732966762205e-06, + "loss": 0.013, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.11793015897274017, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0147, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.14666646718978882, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0159, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1565992683172226, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0152, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.11420148611068726, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0134, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.0945156067609787, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0105, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.0971534475684166, + "learning_rate": 8.95602513567626e-06, + "loss": 0.012, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.12953956425189972, + "learning_rate": 8.952242101985645e-06, + "loss": 0.0144, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.11144538223743439, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0138, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.11536341160535812, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0118, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.08281587064266205, + "learning_rate": 8.940895757786624e-06, + "loss": 0.0091, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.10917636007070541, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0116, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.11539111286401749, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0142, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.09897308051586151, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0141, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.12034150213003159, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0142, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.10506895184516907, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0145, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10428126901388168, + "learning_rate": 8.918215501162718e-06, + "loss": 0.013, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.0966438427567482, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0126, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.09688732028007507, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0136, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10319554805755615, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0133, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11589661985635757, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0136, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.08819710463285446, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0145, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.09449224174022675, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0141, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.11158958822488785, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0124, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.10327963531017303, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0138, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.070926733314991, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0143, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.08468566834926605, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0131, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11933926492929459, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0128, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10370025038719177, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0138, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.13985954225063324, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0144, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.08677119761705399, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0141, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.1023884266614914, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0113, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.14462946355342865, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0132, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.12994524836540222, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0144, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.08624777942895889, + "learning_rate": 8.850274743290876e-06, + "loss": 0.013, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.10459248721599579, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0117, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.12475694715976715, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0131, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09973360598087311, + "learning_rate": 8.838965950000907e-06, + "loss": 0.0142, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.10204580426216125, + "learning_rate": 8.835197287635673e-06, + "loss": 0.0136, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.08137936890125275, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0111, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.09507330507040024, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0128, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.11097851395606995, + "learning_rate": 8.823894110741103e-06, + "loss": 0.0145, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.12435153126716614, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0122, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.1186857596039772, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0123, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10177282989025116, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0119, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.08287174254655838, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0129, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.12975555658340454, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0138, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.07198113948106766, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0116, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.12089285254478455, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0126, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.09319751709699631, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0123, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.11510422080755234, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0147, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.1134580671787262, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0119, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12099597603082657, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0116, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.11170639842748642, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0111, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.0979531928896904, + "learning_rate": 8.774962556249487e-06, + "loss": 0.0137, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.0814109742641449, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0143, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.09910739213228226, + "learning_rate": 8.767441700331522e-06, + "loss": 0.0129, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.10014151781797409, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0125, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.14257559180259705, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.08456768840551376, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0138, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.19197505712509155, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0156, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.10880734771490097, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0143, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.12371200323104858, + "learning_rate": 8.744890501433196e-06, + "loss": 0.0117, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.08023890107870102, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0118, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.10231798142194748, + "learning_rate": 8.737377232363901e-06, + "loss": 0.0107, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.06945646554231644, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0131, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.09935662895441055, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0122, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.15378066897392273, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0142, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.08416267484426498, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0127, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.0862744078040123, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0132, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.11481238156557083, + "learning_rate": 8.71484884879911e-06, + "loss": 0.013, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.09441596269607544, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0126, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08091022074222565, + "learning_rate": 8.707343203095797e-06, + "loss": 0.0126, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.1480294167995453, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0121, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.12529131770133972, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0132, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.12804464995861053, + "learning_rate": 8.696088319396272e-06, + "loss": 0.013, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.10956594347953796, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0129, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.11884522438049316, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0127, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.09215931594371796, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0122, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.1411169320344925, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0139, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.07872501015663147, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0139, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.083979532122612, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0102, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.12711060047149658, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0148, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.10543761402368546, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0131, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.13125181198120117, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0142, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.13816513121128082, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0122, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.09294828027486801, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0126, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.10944795608520508, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0142, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.11008825153112411, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0122, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.10732167959213257, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0138, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.10374412685632706, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0136, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.1021609753370285, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0138, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.10304085910320282, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0125, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.08420020341873169, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0119, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.11911525577306747, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0126, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.07106399536132812, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0117, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.08642835915088654, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0125, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1176094189286232, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0146, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08892391622066498, + "learning_rate": 8.60994458849601e-06, + "loss": 0.012, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.12898282706737518, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0122, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.12732256948947906, + "learning_rate": 8.602465911610909e-06, + "loss": 0.0164, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.1723143309354782, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0146, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.10019178688526154, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0122, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.15125800669193268, + "learning_rate": 8.591251540871672e-06, + "loss": 0.0138, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.08618403971195221, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0106, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.12111509591341019, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0125, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.11654779314994812, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0133, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.12341999262571335, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0136, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.09898588061332703, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0121, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.09825480729341507, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0119, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.09357906877994537, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0125, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13970960676670074, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0131, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.14642572402954102, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0117, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.20111511647701263, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0122, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.12268098443746567, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0127, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.11043815314769745, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0109, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.12570181488990784, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0147, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.1037122830748558, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0134, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.07811714708805084, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0121, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.07688374072313309, + "learning_rate": 8.53151572917108e-06, + "loss": 0.013, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.12285313010215759, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0135, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11350435763597488, + "learning_rate": 8.524057571200438e-06, + "loss": 0.013, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.1195593997836113, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.12681595981121063, + "learning_rate": 8.516601380450119e-06, + "loss": 0.0116, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.10786957293748856, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0117, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.12382174283266068, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0123, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08464337885379791, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0139, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.11292237043380737, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0142, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.13033312559127808, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0125, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.13949616253376007, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0153, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.09846726804971695, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0115, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.12679991126060486, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0149, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.10708379000425339, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0145, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.0848502367734909, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0135, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.13396726548671722, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0142, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13248398900032043, + "learning_rate": 8.471905674372712e-06, + "loss": 0.014, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.10274169594049454, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0139, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.08507993817329407, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0112, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.10394402593374252, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0146, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.10293199867010117, + "learning_rate": 8.457022946052628e-06, + "loss": 0.015, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.16253890097141266, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0128, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.07673029601573944, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0125, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.1329779177904129, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0131, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.0754442885518074, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0125, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.09978003799915314, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0124, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.12813764810562134, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0144, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.08480092883110046, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0124, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.10068657249212265, + "learning_rate": 8.42728135848896e-06, + "loss": 0.012, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.08405598253011703, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0122, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.13025622069835663, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0117, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.11026641726493835, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0127, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.10400493443012238, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0138, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.10374867916107178, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0138, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.128962442278862, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0124, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.07732204347848892, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0124, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.14872954785823822, + "learning_rate": 8.39757171680725e-06, + "loss": 0.013, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.13792090117931366, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0175, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.07409467548131943, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0123, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.11308597028255463, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0149, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.1202300488948822, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0139, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.08481356501579285, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0124, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.07334130257368088, + "learning_rate": 8.375310536902511e-06, + "loss": 0.014, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.09884555637836456, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0137, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.16476134955883026, + "learning_rate": 8.36789416540753e-06, + "loss": 0.0142, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.12900881469249725, + "learning_rate": 8.364186735016254e-06, + "loss": 0.0138, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.10485891997814178, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0146, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.13036216795444489, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0111, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.13155606389045715, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0146, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.11586495488882065, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0135, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.1043778583407402, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0137, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.12916314601898193, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.1257125288248062, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0117, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.20916301012039185, + "learning_rate": 8.334545457592674e-06, + "loss": 0.014, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.13241995871067047, + "learning_rate": 8.330842572848802e-06, + "loss": 0.011, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.1633574515581131, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0141, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.08672885596752167, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0141, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.09803608059883118, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0143, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.13284195959568024, + "learning_rate": 8.316036101469748e-06, + "loss": 0.011, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10581238567829132, + "learning_rate": 8.312335751929666e-06, + "loss": 0.0144, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.11561308056116104, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0111, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.09897869825363159, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0116, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.129640132188797, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0142, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.08553853631019592, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.09117261320352554, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0134, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.1338709443807602, + "learning_rate": 8.290144330085352e-06, + "loss": 0.013, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.11192607134580612, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0122, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.0983581617474556, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0149, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.10853386670351028, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0145, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.09507740288972855, + "learning_rate": 8.275360236495652e-06, + "loss": 0.011, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.08634619414806366, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0146, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.12652689218521118, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0145, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.13687895238399506, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0115, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.11267609149217606, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0139, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.0999043732881546, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0125, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.09523805975914001, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0137, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09847134351730347, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0153, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.0827624723315239, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0093, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.09830755740404129, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0119, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.14614145457744598, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0149, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.1254947930574417, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0129, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.11176569014787674, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0147, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.06597425043582916, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0132, + "step": 80000 + }, + { + "epoch": 23.279022403258654, + "grad_norm": 0.19944901764392853, + "learning_rate": 8.223680373918572e-06, + "loss": 0.0147, + "step": 80010 + }, + { + "epoch": 23.2819319173698, + "grad_norm": 0.09820935130119324, + "learning_rate": 8.219992805682155e-06, + "loss": 0.013, + "step": 80020 + }, + { + "epoch": 23.28484143148094, + "grad_norm": 0.07166675478219986, + "learning_rate": 8.216305752342533e-06, + "loss": 0.0135, + "step": 80030 + }, + { + "epoch": 23.287750945592087, + "grad_norm": 0.08634001016616821, + "learning_rate": 8.21261921417972e-06, + "loss": 0.0128, + "step": 80040 + }, + { + "epoch": 23.29066045970323, + "grad_norm": 0.1467038244009018, + "learning_rate": 8.20893319147369e-06, + "loss": 0.0131, + "step": 80050 + }, + { + "epoch": 23.293569973814375, + "grad_norm": 0.0938076600432396, + "learning_rate": 8.205247684504364e-06, + "loss": 0.0122, + "step": 80060 + }, + { + "epoch": 23.296479487925517, + "grad_norm": 0.06792404502630234, + "learning_rate": 8.20156269355164e-06, + "loss": 0.012, + "step": 80070 + }, + { + "epoch": 23.29938900203666, + "grad_norm": 0.06738271564245224, + "learning_rate": 8.197878218895356e-06, + "loss": 0.0114, + "step": 80080 + }, + { + "epoch": 23.302298516147804, + "grad_norm": 0.16149401664733887, + "learning_rate": 8.194194260815338e-06, + "loss": 0.0121, + "step": 80090 + }, + { + "epoch": 23.305208030258946, + "grad_norm": 0.11988461762666702, + "learning_rate": 8.190510819591348e-06, + "loss": 0.0121, + "step": 80100 + }, + { + "epoch": 23.30811754437009, + "grad_norm": 0.09137140959501266, + "learning_rate": 8.186827895503117e-06, + "loss": 0.0125, + "step": 80110 + }, + { + "epoch": 23.311027058481233, + "grad_norm": 0.0880751833319664, + "learning_rate": 8.183145488830355e-06, + "loss": 0.0116, + "step": 80120 + }, + { + "epoch": 23.313936572592375, + "grad_norm": 0.09596525132656097, + "learning_rate": 8.179463599852705e-06, + "loss": 0.0117, + "step": 80130 + }, + { + "epoch": 23.31684608670352, + "grad_norm": 0.13169355690479279, + "learning_rate": 8.175782228849769e-06, + "loss": 0.0121, + "step": 80140 + }, + { + "epoch": 23.319755600814663, + "grad_norm": 0.1044575497508049, + "learning_rate": 8.17210137610115e-06, + "loss": 0.0132, + "step": 80150 + }, + { + "epoch": 23.322665114925808, + "grad_norm": 0.13589617609977722, + "learning_rate": 8.168421041886373e-06, + "loss": 0.0118, + "step": 80160 + }, + { + "epoch": 23.32557462903695, + "grad_norm": 0.10503333061933517, + "learning_rate": 8.164741226484925e-06, + "loss": 0.0108, + "step": 80170 + }, + { + "epoch": 23.328484143148096, + "grad_norm": 0.09190402179956436, + "learning_rate": 8.161061930176272e-06, + "loss": 0.0129, + "step": 80180 + }, + { + "epoch": 23.331393657259238, + "grad_norm": 0.104253388941288, + "learning_rate": 8.15738315323984e-06, + "loss": 0.0114, + "step": 80190 + }, + { + "epoch": 23.33430317137038, + "grad_norm": 0.10191919654607773, + "learning_rate": 8.153704895955e-06, + "loss": 0.0128, + "step": 80200 + }, + { + "epoch": 23.337212685481525, + "grad_norm": 0.0824175626039505, + "learning_rate": 8.150027158601084e-06, + "loss": 0.0126, + "step": 80210 + }, + { + "epoch": 23.340122199592667, + "grad_norm": 0.08966456353664398, + "learning_rate": 8.146349941457401e-06, + "loss": 0.014, + "step": 80220 + }, + { + "epoch": 23.343031713703812, + "grad_norm": 0.0856788232922554, + "learning_rate": 8.142673244803215e-06, + "loss": 0.01, + "step": 80230 + }, + { + "epoch": 23.345941227814954, + "grad_norm": 0.1027766764163971, + "learning_rate": 8.138997068917736e-06, + "loss": 0.0134, + "step": 80240 + }, + { + "epoch": 23.3488507419261, + "grad_norm": 0.11636431515216827, + "learning_rate": 8.13532141408015e-06, + "loss": 0.0126, + "step": 80250 + }, + { + "epoch": 23.35176025603724, + "grad_norm": 0.09742997586727142, + "learning_rate": 8.131646280569605e-06, + "loss": 0.0117, + "step": 80260 + }, + { + "epoch": 23.354669770148384, + "grad_norm": 0.08142542093992233, + "learning_rate": 8.1279716686652e-06, + "loss": 0.0117, + "step": 80270 + }, + { + "epoch": 23.35757928425953, + "grad_norm": 0.09734860807657242, + "learning_rate": 8.124297578645989e-06, + "loss": 0.0115, + "step": 80280 + }, + { + "epoch": 23.36048879837067, + "grad_norm": 0.10368160158395767, + "learning_rate": 8.120624010791002e-06, + "loss": 0.0142, + "step": 80290 + }, + { + "epoch": 23.363398312481817, + "grad_norm": 0.1619454324245453, + "learning_rate": 8.116950965379225e-06, + "loss": 0.0101, + "step": 80300 + }, + { + "epoch": 23.36630782659296, + "grad_norm": 0.09248818457126617, + "learning_rate": 8.113278442689593e-06, + "loss": 0.0112, + "step": 80310 + }, + { + "epoch": 23.369217340704104, + "grad_norm": 0.11162950843572617, + "learning_rate": 8.109606443001019e-06, + "loss": 0.0126, + "step": 80320 + }, + { + "epoch": 23.372126854815246, + "grad_norm": 0.1364763081073761, + "learning_rate": 8.10593496659236e-06, + "loss": 0.014, + "step": 80330 + }, + { + "epoch": 23.375036368926388, + "grad_norm": 0.113436259329319, + "learning_rate": 8.102264013742442e-06, + "loss": 0.0125, + "step": 80340 + }, + { + "epoch": 23.377945883037533, + "grad_norm": 0.11578390002250671, + "learning_rate": 8.098593584730057e-06, + "loss": 0.0118, + "step": 80350 + }, + { + "epoch": 23.380855397148675, + "grad_norm": 0.11272793263196945, + "learning_rate": 8.09492367983394e-06, + "loss": 0.0135, + "step": 80360 + }, + { + "epoch": 23.38376491125982, + "grad_norm": 0.09545538574457169, + "learning_rate": 8.091254299332803e-06, + "loss": 0.0146, + "step": 80370 + }, + { + "epoch": 23.386674425370963, + "grad_norm": 0.10459403693675995, + "learning_rate": 8.087585443505305e-06, + "loss": 0.013, + "step": 80380 + }, + { + "epoch": 23.389583939482108, + "grad_norm": 0.1124049723148346, + "learning_rate": 8.083917112630078e-06, + "loss": 0.0129, + "step": 80390 + }, + { + "epoch": 23.39249345359325, + "grad_norm": 0.34668007493019104, + "learning_rate": 8.0802493069857e-06, + "loss": 0.012, + "step": 80400 + }, + { + "epoch": 23.395402967704392, + "grad_norm": 0.08212853223085403, + "learning_rate": 8.07658202685072e-06, + "loss": 0.0152, + "step": 80410 + }, + { + "epoch": 23.398312481815537, + "grad_norm": 0.10383888334035873, + "learning_rate": 8.072915272503651e-06, + "loss": 0.0117, + "step": 80420 + }, + { + "epoch": 23.40122199592668, + "grad_norm": 0.12535853683948517, + "learning_rate": 8.069249044222951e-06, + "loss": 0.0152, + "step": 80430 + }, + { + "epoch": 23.404131510037825, + "grad_norm": 0.1187693327665329, + "learning_rate": 8.065583342287044e-06, + "loss": 0.0137, + "step": 80440 + }, + { + "epoch": 23.407041024148967, + "grad_norm": 0.1197381466627121, + "learning_rate": 8.061918166974316e-06, + "loss": 0.0124, + "step": 80450 + }, + { + "epoch": 23.409950538260112, + "grad_norm": 0.09074021875858307, + "learning_rate": 8.058253518563122e-06, + "loss": 0.0145, + "step": 80460 + }, + { + "epoch": 23.412860052371254, + "grad_norm": 0.10440906882286072, + "learning_rate": 8.054589397331757e-06, + "loss": 0.0127, + "step": 80470 + }, + { + "epoch": 23.415769566482396, + "grad_norm": 0.09512275457382202, + "learning_rate": 8.050925803558491e-06, + "loss": 0.0142, + "step": 80480 + }, + { + "epoch": 23.41867908059354, + "grad_norm": 0.06672519445419312, + "learning_rate": 8.047262737521556e-06, + "loss": 0.0119, + "step": 80490 + }, + { + "epoch": 23.421588594704684, + "grad_norm": 0.0820726528763771, + "learning_rate": 8.043600199499131e-06, + "loss": 0.0118, + "step": 80500 + }, + { + "epoch": 23.42449810881583, + "grad_norm": 0.10173840820789337, + "learning_rate": 8.03993818976936e-06, + "loss": 0.0133, + "step": 80510 + }, + { + "epoch": 23.42740762292697, + "grad_norm": 0.06551963835954666, + "learning_rate": 8.03627670861035e-06, + "loss": 0.0111, + "step": 80520 + }, + { + "epoch": 23.430317137038113, + "grad_norm": 0.08851084858179092, + "learning_rate": 8.03261575630017e-06, + "loss": 0.011, + "step": 80530 + }, + { + "epoch": 23.43322665114926, + "grad_norm": 0.11487096548080444, + "learning_rate": 8.028955333116846e-06, + "loss": 0.011, + "step": 80540 + }, + { + "epoch": 23.4361361652604, + "grad_norm": 0.08430162072181702, + "learning_rate": 8.025295439338347e-06, + "loss": 0.0119, + "step": 80550 + }, + { + "epoch": 23.439045679371546, + "grad_norm": 0.06152086704969406, + "learning_rate": 8.021636075242641e-06, + "loss": 0.0122, + "step": 80560 + }, + { + "epoch": 23.441955193482688, + "grad_norm": 0.11052606254816055, + "learning_rate": 8.017977241107622e-06, + "loss": 0.0136, + "step": 80570 + }, + { + "epoch": 23.444864707593833, + "grad_norm": 0.1052686870098114, + "learning_rate": 8.01431893721115e-06, + "loss": 0.012, + "step": 80580 + }, + { + "epoch": 23.447774221704975, + "grad_norm": 0.10118360817432404, + "learning_rate": 8.01066116383105e-06, + "loss": 0.014, + "step": 80590 + }, + { + "epoch": 23.450683735816117, + "grad_norm": 0.10555325448513031, + "learning_rate": 8.00700392124512e-06, + "loss": 0.013, + "step": 80600 + }, + { + "epoch": 23.453593249927263, + "grad_norm": 0.06563732028007507, + "learning_rate": 8.003347209731084e-06, + "loss": 0.0122, + "step": 80610 + }, + { + "epoch": 23.456502764038405, + "grad_norm": 0.1160956397652626, + "learning_rate": 7.999691029566658e-06, + "loss": 0.0125, + "step": 80620 + }, + { + "epoch": 23.45941227814955, + "grad_norm": 0.08643034845590591, + "learning_rate": 7.996035381029494e-06, + "loss": 0.0135, + "step": 80630 + }, + { + "epoch": 23.462321792260692, + "grad_norm": 0.09050534665584564, + "learning_rate": 7.99238026439723e-06, + "loss": 0.0124, + "step": 80640 + }, + { + "epoch": 23.465231306371837, + "grad_norm": 0.07935993373394012, + "learning_rate": 7.988725679947432e-06, + "loss": 0.0129, + "step": 80650 + }, + { + "epoch": 23.46814082048298, + "grad_norm": 0.11728908121585846, + "learning_rate": 7.985071627957645e-06, + "loss": 0.0122, + "step": 80660 + }, + { + "epoch": 23.47105033459412, + "grad_norm": 0.18306072056293488, + "learning_rate": 7.981418108705381e-06, + "loss": 0.0131, + "step": 80670 + }, + { + "epoch": 23.473959848705267, + "grad_norm": 0.10635816305875778, + "learning_rate": 7.977765122468089e-06, + "loss": 0.0119, + "step": 80680 + }, + { + "epoch": 23.47686936281641, + "grad_norm": 0.13399194180965424, + "learning_rate": 7.974112669523197e-06, + "loss": 0.0164, + "step": 80690 + }, + { + "epoch": 23.479778876927554, + "grad_norm": 0.12431732565164566, + "learning_rate": 7.970460750148077e-06, + "loss": 0.0112, + "step": 80700 + }, + { + "epoch": 23.482688391038696, + "grad_norm": 0.12431129813194275, + "learning_rate": 7.96680936462007e-06, + "loss": 0.0143, + "step": 80710 + }, + { + "epoch": 23.48559790514984, + "grad_norm": 0.09648348391056061, + "learning_rate": 7.963158513216482e-06, + "loss": 0.0105, + "step": 80720 + }, + { + "epoch": 23.488507419260984, + "grad_norm": 0.11739730834960938, + "learning_rate": 7.959508196214568e-06, + "loss": 0.0131, + "step": 80730 + }, + { + "epoch": 23.491416933372125, + "grad_norm": 0.109571672976017, + "learning_rate": 7.955858413891536e-06, + "loss": 0.0124, + "step": 80740 + }, + { + "epoch": 23.49432644748327, + "grad_norm": 0.09891501069068909, + "learning_rate": 7.952209166524571e-06, + "loss": 0.0123, + "step": 80750 + }, + { + "epoch": 23.497235961594413, + "grad_norm": 0.11239849030971527, + "learning_rate": 7.948560454390814e-06, + "loss": 0.0114, + "step": 80760 + }, + { + "epoch": 23.50014547570556, + "grad_norm": 0.08035823702812195, + "learning_rate": 7.94491227776735e-06, + "loss": 0.0106, + "step": 80770 + }, + { + "epoch": 23.5030549898167, + "grad_norm": 0.1137743890285492, + "learning_rate": 7.941264636931237e-06, + "loss": 0.0125, + "step": 80780 + }, + { + "epoch": 23.505964503927846, + "grad_norm": 0.11350683122873306, + "learning_rate": 7.9376175321595e-06, + "loss": 0.0119, + "step": 80790 + }, + { + "epoch": 23.508874018038988, + "grad_norm": 0.15854588150978088, + "learning_rate": 7.933970963729102e-06, + "loss": 0.0111, + "step": 80800 + }, + { + "epoch": 23.51178353215013, + "grad_norm": 0.09225080907344818, + "learning_rate": 7.93032493191697e-06, + "loss": 0.0145, + "step": 80810 + }, + { + "epoch": 23.514693046261275, + "grad_norm": 0.07517179846763611, + "learning_rate": 7.926679437000006e-06, + "loss": 0.0129, + "step": 80820 + }, + { + "epoch": 23.517602560372417, + "grad_norm": 0.09395122528076172, + "learning_rate": 7.923034479255065e-06, + "loss": 0.014, + "step": 80830 + }, + { + "epoch": 23.520512074483563, + "grad_norm": 0.0953364223241806, + "learning_rate": 7.91939005895895e-06, + "loss": 0.0132, + "step": 80840 + }, + { + "epoch": 23.523421588594704, + "grad_norm": 0.1388244777917862, + "learning_rate": 7.91574617638842e-06, + "loss": 0.0142, + "step": 80850 + }, + { + "epoch": 23.526331102705846, + "grad_norm": 0.10685098171234131, + "learning_rate": 7.912102831820231e-06, + "loss": 0.0143, + "step": 80860 + }, + { + "epoch": 23.529240616816992, + "grad_norm": 0.09484872967004776, + "learning_rate": 7.908460025531051e-06, + "loss": 0.0138, + "step": 80870 + }, + { + "epoch": 23.532150130928134, + "grad_norm": 0.0918048843741417, + "learning_rate": 7.90481775779753e-06, + "loss": 0.0136, + "step": 80880 + }, + { + "epoch": 23.53505964503928, + "grad_norm": 0.15061822533607483, + "learning_rate": 7.901176028896274e-06, + "loss": 0.0149, + "step": 80890 + }, + { + "epoch": 23.53796915915042, + "grad_norm": 0.0951741561293602, + "learning_rate": 7.897534839103856e-06, + "loss": 0.0118, + "step": 80900 + }, + { + "epoch": 23.540878673261567, + "grad_norm": 0.08154623210430145, + "learning_rate": 7.893894188696797e-06, + "loss": 0.0144, + "step": 80910 + }, + { + "epoch": 23.54378818737271, + "grad_norm": 0.10864793509244919, + "learning_rate": 7.89025407795157e-06, + "loss": 0.0132, + "step": 80920 + }, + { + "epoch": 23.54669770148385, + "grad_norm": 0.12371568381786346, + "learning_rate": 7.886614507144628e-06, + "loss": 0.0128, + "step": 80930 + }, + { + "epoch": 23.549607215594996, + "grad_norm": 0.13036499917507172, + "learning_rate": 7.882975476552372e-06, + "loss": 0.0135, + "step": 80940 + }, + { + "epoch": 23.552516729706138, + "grad_norm": 0.10305178165435791, + "learning_rate": 7.87933698645116e-06, + "loss": 0.0104, + "step": 80950 + }, + { + "epoch": 23.555426243817283, + "grad_norm": 0.09468064457178116, + "learning_rate": 7.875699037117306e-06, + "loss": 0.0131, + "step": 80960 + }, + { + "epoch": 23.558335757928425, + "grad_norm": 0.12249180674552917, + "learning_rate": 7.872061628827103e-06, + "loss": 0.0132, + "step": 80970 + }, + { + "epoch": 23.56124527203957, + "grad_norm": 0.0880570113658905, + "learning_rate": 7.86842476185677e-06, + "loss": 0.0103, + "step": 80980 + }, + { + "epoch": 23.564154786150713, + "grad_norm": 0.12612831592559814, + "learning_rate": 7.864788436482522e-06, + "loss": 0.0123, + "step": 80990 + }, + { + "epoch": 23.567064300261855, + "grad_norm": 0.12270201742649078, + "learning_rate": 7.861152652980497e-06, + "loss": 0.0121, + "step": 81000 + }, + { + "epoch": 23.569973814373, + "grad_norm": 0.10333582013845444, + "learning_rate": 7.85751741162682e-06, + "loss": 0.0098, + "step": 81010 + }, + { + "epoch": 23.572883328484142, + "grad_norm": 0.15205809473991394, + "learning_rate": 7.853882712697554e-06, + "loss": 0.0144, + "step": 81020 + }, + { + "epoch": 23.575792842595288, + "grad_norm": 0.09015808999538422, + "learning_rate": 7.850248556468741e-06, + "loss": 0.0109, + "step": 81030 + }, + { + "epoch": 23.57870235670643, + "grad_norm": 0.11366626620292664, + "learning_rate": 7.846614943216361e-06, + "loss": 0.0121, + "step": 81040 + }, + { + "epoch": 23.581611870817575, + "grad_norm": 0.08618906885385513, + "learning_rate": 7.842981873216368e-06, + "loss": 0.0137, + "step": 81050 + }, + { + "epoch": 23.584521384928717, + "grad_norm": 0.11496937274932861, + "learning_rate": 7.839349346744675e-06, + "loss": 0.0144, + "step": 81060 + }, + { + "epoch": 23.58743089903986, + "grad_norm": 0.11915834993124008, + "learning_rate": 7.835717364077138e-06, + "loss": 0.014, + "step": 81070 + }, + { + "epoch": 23.590340413151004, + "grad_norm": 0.12172401696443558, + "learning_rate": 7.832085925489584e-06, + "loss": 0.0125, + "step": 81080 + }, + { + "epoch": 23.593249927262146, + "grad_norm": 0.10628852248191833, + "learning_rate": 7.82845503125781e-06, + "loss": 0.0135, + "step": 81090 + }, + { + "epoch": 23.596159441373292, + "grad_norm": 0.11503509432077408, + "learning_rate": 7.824824681657544e-06, + "loss": 0.0128, + "step": 81100 + }, + { + "epoch": 23.599068955484434, + "grad_norm": 0.11923188716173172, + "learning_rate": 7.821194876964489e-06, + "loss": 0.0132, + "step": 81110 + }, + { + "epoch": 23.60197846959558, + "grad_norm": 0.09143807739019394, + "learning_rate": 7.817565617454305e-06, + "loss": 0.0131, + "step": 81120 + }, + { + "epoch": 23.60488798370672, + "grad_norm": 0.07396473735570908, + "learning_rate": 7.813936903402619e-06, + "loss": 0.0122, + "step": 81130 + }, + { + "epoch": 23.607797497817863, + "grad_norm": 0.11674617975950241, + "learning_rate": 7.810308735085001e-06, + "loss": 0.0127, + "step": 81140 + }, + { + "epoch": 23.61070701192901, + "grad_norm": 0.12394711375236511, + "learning_rate": 7.806681112776976e-06, + "loss": 0.0127, + "step": 81150 + }, + { + "epoch": 23.61361652604015, + "grad_norm": 0.14487721025943756, + "learning_rate": 7.803054036754058e-06, + "loss": 0.0124, + "step": 81160 + }, + { + "epoch": 23.616526040151296, + "grad_norm": 0.1078348383307457, + "learning_rate": 7.799427507291694e-06, + "loss": 0.0143, + "step": 81170 + }, + { + "epoch": 23.619435554262438, + "grad_norm": 0.10270193219184875, + "learning_rate": 7.795801524665284e-06, + "loss": 0.0115, + "step": 81180 + }, + { + "epoch": 23.622345068373583, + "grad_norm": 0.06343881785869598, + "learning_rate": 7.792176089150203e-06, + "loss": 0.0129, + "step": 81190 + }, + { + "epoch": 23.625254582484725, + "grad_norm": 0.08612993359565735, + "learning_rate": 7.788551201021787e-06, + "loss": 0.0116, + "step": 81200 + }, + { + "epoch": 23.628164096595867, + "grad_norm": 0.1368905007839203, + "learning_rate": 7.784926860555315e-06, + "loss": 0.014, + "step": 81210 + }, + { + "epoch": 23.631073610707013, + "grad_norm": 0.11004527658224106, + "learning_rate": 7.781303068026029e-06, + "loss": 0.0118, + "step": 81220 + }, + { + "epoch": 23.633983124818155, + "grad_norm": 0.11310701817274094, + "learning_rate": 7.777679823709134e-06, + "loss": 0.0149, + "step": 81230 + }, + { + "epoch": 23.6368926389293, + "grad_norm": 0.09327834844589233, + "learning_rate": 7.7740571278798e-06, + "loss": 0.0136, + "step": 81240 + }, + { + "epoch": 23.639802153040442, + "grad_norm": 0.15093250572681427, + "learning_rate": 7.770434980813133e-06, + "loss": 0.0133, + "step": 81250 + }, + { + "epoch": 23.642711667151584, + "grad_norm": 0.09899720549583435, + "learning_rate": 7.76681338278422e-06, + "loss": 0.0117, + "step": 81260 + }, + { + "epoch": 23.64562118126273, + "grad_norm": 0.11601489782333374, + "learning_rate": 7.7631923340681e-06, + "loss": 0.0128, + "step": 81270 + }, + { + "epoch": 23.64853069537387, + "grad_norm": 0.09682130813598633, + "learning_rate": 7.759571834939765e-06, + "loss": 0.0121, + "step": 81280 + }, + { + "epoch": 23.651440209485017, + "grad_norm": 0.10942123830318451, + "learning_rate": 7.75595188567416e-06, + "loss": 0.0117, + "step": 81290 + }, + { + "epoch": 23.65434972359616, + "grad_norm": 0.08153918385505676, + "learning_rate": 7.7523324865462e-06, + "loss": 0.0131, + "step": 81300 + }, + { + "epoch": 23.657259237707304, + "grad_norm": 0.08599690347909927, + "learning_rate": 7.748713637830768e-06, + "loss": 0.0115, + "step": 81310 + }, + { + "epoch": 23.660168751818446, + "grad_norm": 0.08310934901237488, + "learning_rate": 7.745095339802672e-06, + "loss": 0.0127, + "step": 81320 + }, + { + "epoch": 23.663078265929588, + "grad_norm": 0.10698913782835007, + "learning_rate": 7.741477592736712e-06, + "loss": 0.0126, + "step": 81330 + }, + { + "epoch": 23.665987780040734, + "grad_norm": 0.14655916392803192, + "learning_rate": 7.737860396907623e-06, + "loss": 0.014, + "step": 81340 + }, + { + "epoch": 23.668897294151876, + "grad_norm": 0.07894794642925262, + "learning_rate": 7.73424375259011e-06, + "loss": 0.0116, + "step": 81350 + }, + { + "epoch": 23.67180680826302, + "grad_norm": 0.09374310821294785, + "learning_rate": 7.730627660058841e-06, + "loss": 0.0101, + "step": 81360 + }, + { + "epoch": 23.674716322374163, + "grad_norm": 0.1619149148464203, + "learning_rate": 7.72701211958842e-06, + "loss": 0.0131, + "step": 81370 + }, + { + "epoch": 23.67762583648531, + "grad_norm": 0.0957368016242981, + "learning_rate": 7.723397131453438e-06, + "loss": 0.0119, + "step": 81380 + }, + { + "epoch": 23.68053535059645, + "grad_norm": 0.10419198870658875, + "learning_rate": 7.719782695928416e-06, + "loss": 0.0126, + "step": 81390 + }, + { + "epoch": 23.683444864707592, + "grad_norm": 0.1281127780675888, + "learning_rate": 7.716168813287858e-06, + "loss": 0.0142, + "step": 81400 + }, + { + "epoch": 23.686354378818738, + "grad_norm": 0.10857229679822922, + "learning_rate": 7.712555483806205e-06, + "loss": 0.0141, + "step": 81410 + }, + { + "epoch": 23.68926389292988, + "grad_norm": 0.10917466878890991, + "learning_rate": 7.70894270775787e-06, + "loss": 0.0127, + "step": 81420 + }, + { + "epoch": 23.692173407041025, + "grad_norm": 0.09018818289041519, + "learning_rate": 7.705330485417224e-06, + "loss": 0.0126, + "step": 81430 + }, + { + "epoch": 23.695082921152167, + "grad_norm": 0.15100347995758057, + "learning_rate": 7.70171881705859e-06, + "loss": 0.0146, + "step": 81440 + }, + { + "epoch": 23.697992435263313, + "grad_norm": 0.08587733656167984, + "learning_rate": 7.698107702956232e-06, + "loss": 0.0129, + "step": 81450 + }, + { + "epoch": 23.700901949374455, + "grad_norm": 0.1320592761039734, + "learning_rate": 7.69449714338442e-06, + "loss": 0.0123, + "step": 81460 + }, + { + "epoch": 23.703811463485597, + "grad_norm": 0.1478150337934494, + "learning_rate": 7.690887138617335e-06, + "loss": 0.0133, + "step": 81470 + }, + { + "epoch": 23.706720977596742, + "grad_norm": 0.11000087857246399, + "learning_rate": 7.687277688929134e-06, + "loss": 0.0113, + "step": 81480 + }, + { + "epoch": 23.709630491707884, + "grad_norm": 0.11071797460317612, + "learning_rate": 7.68366879459393e-06, + "loss": 0.0128, + "step": 81490 + }, + { + "epoch": 23.71254000581903, + "grad_norm": 0.10804122686386108, + "learning_rate": 7.680060455885804e-06, + "loss": 0.0136, + "step": 81500 + }, + { + "epoch": 23.71544951993017, + "grad_norm": 0.11369496583938599, + "learning_rate": 7.676452673078779e-06, + "loss": 0.0118, + "step": 81510 + }, + { + "epoch": 23.718359034041313, + "grad_norm": 0.09902388602495193, + "learning_rate": 7.672845446446835e-06, + "loss": 0.0118, + "step": 81520 + }, + { + "epoch": 23.72126854815246, + "grad_norm": 0.1289895921945572, + "learning_rate": 7.669238776263927e-06, + "loss": 0.0134, + "step": 81530 + }, + { + "epoch": 23.7241780622636, + "grad_norm": 0.13489077985286713, + "learning_rate": 7.665632662803961e-06, + "loss": 0.0115, + "step": 81540 + }, + { + "epoch": 23.727087576374746, + "grad_norm": 0.06097550690174103, + "learning_rate": 7.662027106340785e-06, + "loss": 0.0136, + "step": 81550 + }, + { + "epoch": 23.729997090485888, + "grad_norm": 0.09678100794553757, + "learning_rate": 7.658422107148222e-06, + "loss": 0.0098, + "step": 81560 + }, + { + "epoch": 23.732906604597034, + "grad_norm": 0.09136251360177994, + "learning_rate": 7.654817665500058e-06, + "loss": 0.013, + "step": 81570 + }, + { + "epoch": 23.735816118708176, + "grad_norm": 0.09218855947256088, + "learning_rate": 7.651213781670016e-06, + "loss": 0.0127, + "step": 81580 + }, + { + "epoch": 23.738725632819317, + "grad_norm": 0.07712986320257187, + "learning_rate": 7.647610455931784e-06, + "loss": 0.0123, + "step": 81590 + }, + { + "epoch": 23.741635146930463, + "grad_norm": 0.0797911211848259, + "learning_rate": 7.644007688559018e-06, + "loss": 0.0113, + "step": 81600 + }, + { + "epoch": 23.744544661041605, + "grad_norm": 0.09198153764009476, + "learning_rate": 7.64040547982533e-06, + "loss": 0.0126, + "step": 81610 + }, + { + "epoch": 23.74745417515275, + "grad_norm": 0.11342693865299225, + "learning_rate": 7.636803830004267e-06, + "loss": 0.0126, + "step": 81620 + }, + { + "epoch": 23.750363689263892, + "grad_norm": 0.08946854621171951, + "learning_rate": 7.633202739369368e-06, + "loss": 0.0107, + "step": 81630 + }, + { + "epoch": 23.753273203375038, + "grad_norm": 0.10234414041042328, + "learning_rate": 7.6296022081941e-06, + "loss": 0.0114, + "step": 81640 + }, + { + "epoch": 23.75618271748618, + "grad_norm": 0.1319761723279953, + "learning_rate": 7.6260022367519065e-06, + "loss": 0.0124, + "step": 81650 + }, + { + "epoch": 23.75909223159732, + "grad_norm": 0.11398373544216156, + "learning_rate": 7.622402825316175e-06, + "loss": 0.0125, + "step": 81660 + }, + { + "epoch": 23.762001745708467, + "grad_norm": 0.10755864530801773, + "learning_rate": 7.61880397416026e-06, + "loss": 0.0124, + "step": 81670 + }, + { + "epoch": 23.76491125981961, + "grad_norm": 0.12102973461151123, + "learning_rate": 7.61520568355748e-06, + "loss": 0.0115, + "step": 81680 + }, + { + "epoch": 23.767820773930755, + "grad_norm": 0.14041413366794586, + "learning_rate": 7.611607953781084e-06, + "loss": 0.0131, + "step": 81690 + }, + { + "epoch": 23.770730288041896, + "grad_norm": 0.11478327959775925, + "learning_rate": 7.6080107851043116e-06, + "loss": 0.0127, + "step": 81700 + }, + { + "epoch": 23.773639802153042, + "grad_norm": 0.14033357799053192, + "learning_rate": 7.6044141778003314e-06, + "loss": 0.015, + "step": 81710 + }, + { + "epoch": 23.776549316264184, + "grad_norm": 0.11536005139350891, + "learning_rate": 7.600818132142289e-06, + "loss": 0.0141, + "step": 81720 + }, + { + "epoch": 23.779458830375326, + "grad_norm": 0.10734536498785019, + "learning_rate": 7.597222648403283e-06, + "loss": 0.0117, + "step": 81730 + }, + { + "epoch": 23.78236834448647, + "grad_norm": 0.0891343355178833, + "learning_rate": 7.593627726856365e-06, + "loss": 0.0117, + "step": 81740 + }, + { + "epoch": 23.785277858597613, + "grad_norm": 0.09240922331809998, + "learning_rate": 7.5900333677745335e-06, + "loss": 0.0104, + "step": 81750 + }, + { + "epoch": 23.78818737270876, + "grad_norm": 0.10601887851953506, + "learning_rate": 7.586439571430771e-06, + "loss": 0.0137, + "step": 81760 + }, + { + "epoch": 23.7910968868199, + "grad_norm": 0.12168681621551514, + "learning_rate": 7.5828463380980015e-06, + "loss": 0.014, + "step": 81770 + }, + { + "epoch": 23.794006400931046, + "grad_norm": 0.11172056198120117, + "learning_rate": 7.579253668049098e-06, + "loss": 0.0145, + "step": 81780 + }, + { + "epoch": 23.796915915042188, + "grad_norm": 0.1011413112282753, + "learning_rate": 7.575661561556905e-06, + "loss": 0.0116, + "step": 81790 + }, + { + "epoch": 23.79982542915333, + "grad_norm": 0.11835134774446487, + "learning_rate": 7.572070018894225e-06, + "loss": 0.0137, + "step": 81800 + }, + { + "epoch": 23.802734943264475, + "grad_norm": 0.1536318063735962, + "learning_rate": 7.5684790403338096e-06, + "loss": 0.012, + "step": 81810 + }, + { + "epoch": 23.805644457375617, + "grad_norm": 0.0950169786810875, + "learning_rate": 7.564888626148361e-06, + "loss": 0.0118, + "step": 81820 + }, + { + "epoch": 23.808553971486763, + "grad_norm": 0.11085739731788635, + "learning_rate": 7.561298776610552e-06, + "loss": 0.0124, + "step": 81830 + }, + { + "epoch": 23.811463485597905, + "grad_norm": 0.11517464369535446, + "learning_rate": 7.557709491993017e-06, + "loss": 0.0146, + "step": 81840 + }, + { + "epoch": 23.81437299970905, + "grad_norm": 0.14310887455940247, + "learning_rate": 7.55412077256833e-06, + "loss": 0.0141, + "step": 81850 + }, + { + "epoch": 23.817282513820192, + "grad_norm": 0.0991383045911789, + "learning_rate": 7.550532618609022e-06, + "loss": 0.0119, + "step": 81860 + }, + { + "epoch": 23.820192027931334, + "grad_norm": 0.08636868745088577, + "learning_rate": 7.54694503038761e-06, + "loss": 0.0129, + "step": 81870 + }, + { + "epoch": 23.82310154204248, + "grad_norm": 0.11465732753276825, + "learning_rate": 7.543358008176537e-06, + "loss": 0.0117, + "step": 81880 + }, + { + "epoch": 23.82601105615362, + "grad_norm": 0.15294228494167328, + "learning_rate": 7.539771552248207e-06, + "loss": 0.0104, + "step": 81890 + }, + { + "epoch": 23.828920570264767, + "grad_norm": 0.11894837021827698, + "learning_rate": 7.536185662874997e-06, + "loss": 0.0142, + "step": 81900 + }, + { + "epoch": 23.83183008437591, + "grad_norm": 0.11275901645421982, + "learning_rate": 7.532600340329232e-06, + "loss": 0.0119, + "step": 81910 + }, + { + "epoch": 23.834739598487054, + "grad_norm": 0.08029196411371231, + "learning_rate": 7.529015584883188e-06, + "loss": 0.0116, + "step": 81920 + }, + { + "epoch": 23.837649112598196, + "grad_norm": 0.11938098818063736, + "learning_rate": 7.5254313968091095e-06, + "loss": 0.0127, + "step": 81930 + }, + { + "epoch": 23.84055862670934, + "grad_norm": 0.09227802604436874, + "learning_rate": 7.521847776379183e-06, + "loss": 0.0134, + "step": 81940 + }, + { + "epoch": 23.843468140820484, + "grad_norm": 0.10166507214307785, + "learning_rate": 7.518264723865572e-06, + "loss": 0.0124, + "step": 81950 + }, + { + "epoch": 23.846377654931626, + "grad_norm": 0.13323524594306946, + "learning_rate": 7.514682239540374e-06, + "loss": 0.0135, + "step": 81960 + }, + { + "epoch": 23.84928716904277, + "grad_norm": 0.18477675318717957, + "learning_rate": 7.5111003236756625e-06, + "loss": 0.0141, + "step": 81970 + }, + { + "epoch": 23.852196683153913, + "grad_norm": 0.11462473124265671, + "learning_rate": 7.507518976543463e-06, + "loss": 0.0121, + "step": 81980 + }, + { + "epoch": 23.855106197265055, + "grad_norm": 0.14673922955989838, + "learning_rate": 7.503938198415746e-06, + "loss": 0.0131, + "step": 81990 + }, + { + "epoch": 23.8580157113762, + "grad_norm": 0.11723819375038147, + "learning_rate": 7.5003579895644596e-06, + "loss": 0.015, + "step": 82000 + }, + { + "epoch": 23.860925225487343, + "grad_norm": 0.1134917214512825, + "learning_rate": 7.496778350261485e-06, + "loss": 0.0135, + "step": 82010 + }, + { + "epoch": 23.863834739598488, + "grad_norm": 0.12128591537475586, + "learning_rate": 7.493199280778683e-06, + "loss": 0.013, + "step": 82020 + }, + { + "epoch": 23.86674425370963, + "grad_norm": 0.13865070044994354, + "learning_rate": 7.489620781387849e-06, + "loss": 0.0142, + "step": 82030 + }, + { + "epoch": 23.869653767820775, + "grad_norm": 0.0896529033780098, + "learning_rate": 7.486042852360758e-06, + "loss": 0.0109, + "step": 82040 + }, + { + "epoch": 23.872563281931917, + "grad_norm": 0.09412191063165665, + "learning_rate": 7.482465493969121e-06, + "loss": 0.0138, + "step": 82050 + }, + { + "epoch": 23.87547279604306, + "grad_norm": 0.1172008290886879, + "learning_rate": 7.4788887064846155e-06, + "loss": 0.0127, + "step": 82060 + }, + { + "epoch": 23.878382310154205, + "grad_norm": 0.1011468768119812, + "learning_rate": 7.4753124901788845e-06, + "loss": 0.012, + "step": 82070 + }, + { + "epoch": 23.881291824265347, + "grad_norm": 0.10312607884407043, + "learning_rate": 7.471736845323507e-06, + "loss": 0.0135, + "step": 82080 + }, + { + "epoch": 23.884201338376492, + "grad_norm": 0.08789291232824326, + "learning_rate": 7.4681617721900324e-06, + "loss": 0.0141, + "step": 82090 + }, + { + "epoch": 23.887110852487634, + "grad_norm": 0.08846236020326614, + "learning_rate": 7.4645872710499705e-06, + "loss": 0.0098, + "step": 82100 + }, + { + "epoch": 23.89002036659878, + "grad_norm": 0.1339399814605713, + "learning_rate": 7.4610133421747765e-06, + "loss": 0.0142, + "step": 82110 + }, + { + "epoch": 23.89292988070992, + "grad_norm": 0.13560190796852112, + "learning_rate": 7.457439985835862e-06, + "loss": 0.0133, + "step": 82120 + }, + { + "epoch": 23.895839394821063, + "grad_norm": 0.08713413774967194, + "learning_rate": 7.453867202304602e-06, + "loss": 0.0122, + "step": 82130 + }, + { + "epoch": 23.89874890893221, + "grad_norm": 0.10207775980234146, + "learning_rate": 7.450294991852334e-06, + "loss": 0.014, + "step": 82140 + }, + { + "epoch": 23.90165842304335, + "grad_norm": 0.10218358039855957, + "learning_rate": 7.446723354750337e-06, + "loss": 0.0114, + "step": 82150 + }, + { + "epoch": 23.904567937154496, + "grad_norm": 0.11706096678972244, + "learning_rate": 7.443152291269844e-06, + "loss": 0.0118, + "step": 82160 + }, + { + "epoch": 23.90747745126564, + "grad_norm": 0.10717874765396118, + "learning_rate": 7.439581801682075e-06, + "loss": 0.0139, + "step": 82170 + }, + { + "epoch": 23.910386965376784, + "grad_norm": 0.09272971004247665, + "learning_rate": 7.436011886258174e-06, + "loss": 0.0118, + "step": 82180 + }, + { + "epoch": 23.913296479487926, + "grad_norm": 0.08789035677909851, + "learning_rate": 7.4324425452692445e-06, + "loss": 0.0119, + "step": 82190 + }, + { + "epoch": 23.916205993599068, + "grad_norm": 0.10102546960115433, + "learning_rate": 7.428873778986364e-06, + "loss": 0.0127, + "step": 82200 + }, + { + "epoch": 23.919115507710213, + "grad_norm": 0.14371797442436218, + "learning_rate": 7.42530558768056e-06, + "loss": 0.012, + "step": 82210 + }, + { + "epoch": 23.922025021821355, + "grad_norm": 0.09243465960025787, + "learning_rate": 7.4217379716228075e-06, + "loss": 0.0129, + "step": 82220 + }, + { + "epoch": 23.9249345359325, + "grad_norm": 0.13050049543380737, + "learning_rate": 7.418170931084039e-06, + "loss": 0.0116, + "step": 82230 + }, + { + "epoch": 23.927844050043642, + "grad_norm": 0.13714571297168732, + "learning_rate": 7.414604466335152e-06, + "loss": 0.0147, + "step": 82240 + }, + { + "epoch": 23.930753564154784, + "grad_norm": 0.131483793258667, + "learning_rate": 7.411038577647002e-06, + "loss": 0.012, + "step": 82250 + }, + { + "epoch": 23.93366307826593, + "grad_norm": 0.1075965166091919, + "learning_rate": 7.407473265290385e-06, + "loss": 0.014, + "step": 82260 + }, + { + "epoch": 23.936572592377072, + "grad_norm": 0.08791981637477875, + "learning_rate": 7.403908529536066e-06, + "loss": 0.0106, + "step": 82270 + }, + { + "epoch": 23.939482106488217, + "grad_norm": 0.14243891835212708, + "learning_rate": 7.40034437065477e-06, + "loss": 0.0125, + "step": 82280 + }, + { + "epoch": 23.94239162059936, + "grad_norm": 0.11222469806671143, + "learning_rate": 7.396780788917161e-06, + "loss": 0.0126, + "step": 82290 + }, + { + "epoch": 23.945301134710505, + "grad_norm": 0.1121295914053917, + "learning_rate": 7.393217784593878e-06, + "loss": 0.0131, + "step": 82300 + }, + { + "epoch": 23.948210648821647, + "grad_norm": 0.09923163801431656, + "learning_rate": 7.389655357955499e-06, + "loss": 0.0118, + "step": 82310 + }, + { + "epoch": 23.95112016293279, + "grad_norm": 0.14223536849021912, + "learning_rate": 7.386093509272578e-06, + "loss": 0.0127, + "step": 82320 + }, + { + "epoch": 23.954029677043934, + "grad_norm": 0.11418648064136505, + "learning_rate": 7.3825322388156016e-06, + "loss": 0.0128, + "step": 82330 + }, + { + "epoch": 23.956939191155076, + "grad_norm": 0.13791611790657043, + "learning_rate": 7.3789715468550365e-06, + "loss": 0.0142, + "step": 82340 + }, + { + "epoch": 23.95984870526622, + "grad_norm": 0.10311667621135712, + "learning_rate": 7.375411433661281e-06, + "loss": 0.0128, + "step": 82350 + }, + { + "epoch": 23.962758219377363, + "grad_norm": 0.1287662386894226, + "learning_rate": 7.371851899504712e-06, + "loss": 0.012, + "step": 82360 + }, + { + "epoch": 23.96566773348851, + "grad_norm": 0.09995871037244797, + "learning_rate": 7.368292944655653e-06, + "loss": 0.0133, + "step": 82370 + }, + { + "epoch": 23.96857724759965, + "grad_norm": 0.07809341698884964, + "learning_rate": 7.3647345693843746e-06, + "loss": 0.0122, + "step": 82380 + }, + { + "epoch": 23.971486761710793, + "grad_norm": 0.11243429780006409, + "learning_rate": 7.361176773961118e-06, + "loss": 0.0117, + "step": 82390 + }, + { + "epoch": 23.974396275821938, + "grad_norm": 0.16213670372962952, + "learning_rate": 7.357619558656079e-06, + "loss": 0.012, + "step": 82400 + }, + { + "epoch": 23.97730578993308, + "grad_norm": 0.09824518859386444, + "learning_rate": 7.3540629237394e-06, + "loss": 0.0154, + "step": 82410 + }, + { + "epoch": 23.980215304044226, + "grad_norm": 0.17611005902290344, + "learning_rate": 7.350506869481177e-06, + "loss": 0.0122, + "step": 82420 + }, + { + "epoch": 23.983124818155368, + "grad_norm": 0.13222116231918335, + "learning_rate": 7.3469513961514745e-06, + "loss": 0.0128, + "step": 82430 + }, + { + "epoch": 23.986034332266513, + "grad_norm": 0.07638631016016006, + "learning_rate": 7.343396504020313e-06, + "loss": 0.0112, + "step": 82440 + }, + { + "epoch": 23.988943846377655, + "grad_norm": 0.15621493756771088, + "learning_rate": 7.339842193357659e-06, + "loss": 0.0142, + "step": 82450 + }, + { + "epoch": 23.991853360488797, + "grad_norm": 0.08129988610744476, + "learning_rate": 7.336288464433427e-06, + "loss": 0.0116, + "step": 82460 + }, + { + "epoch": 23.994762874599942, + "grad_norm": 0.09794914722442627, + "learning_rate": 7.332735317517519e-06, + "loss": 0.0122, + "step": 82470 + }, + { + "epoch": 23.997672388711084, + "grad_norm": 0.10776340961456299, + "learning_rate": 7.329182752879765e-06, + "loss": 0.0131, + "step": 82480 + }, + { + "epoch": 24.00058190282223, + "grad_norm": 0.14134608209133148, + "learning_rate": 7.3256307707899525e-06, + "loss": 0.0125, + "step": 82490 + }, + { + "epoch": 24.00349141693337, + "grad_norm": 0.11423341929912567, + "learning_rate": 7.322079371517838e-06, + "loss": 0.0133, + "step": 82500 + }, + { + "epoch": 24.006400931044517, + "grad_norm": 0.09195499122142792, + "learning_rate": 7.31852855533313e-06, + "loss": 0.0123, + "step": 82510 + }, + { + "epoch": 24.00931044515566, + "grad_norm": 0.12623442709445953, + "learning_rate": 7.314978322505485e-06, + "loss": 0.0122, + "step": 82520 + }, + { + "epoch": 24.0122199592668, + "grad_norm": 0.09605465084314346, + "learning_rate": 7.311428673304513e-06, + "loss": 0.0122, + "step": 82530 + }, + { + "epoch": 24.015129473377947, + "grad_norm": 0.1834714114665985, + "learning_rate": 7.307879607999794e-06, + "loss": 0.0136, + "step": 82540 + }, + { + "epoch": 24.01803898748909, + "grad_norm": 0.10073646157979965, + "learning_rate": 7.304331126860859e-06, + "loss": 0.0131, + "step": 82550 + }, + { + "epoch": 24.020948501600234, + "grad_norm": 0.09505777060985565, + "learning_rate": 7.300783230157185e-06, + "loss": 0.0119, + "step": 82560 + }, + { + "epoch": 24.023858015711376, + "grad_norm": 0.11527670174837112, + "learning_rate": 7.297235918158213e-06, + "loss": 0.0128, + "step": 82570 + }, + { + "epoch": 24.02676752982252, + "grad_norm": 0.10394103080034256, + "learning_rate": 7.293689191133345e-06, + "loss": 0.0122, + "step": 82580 + }, + { + "epoch": 24.029677043933663, + "grad_norm": 0.11201265454292297, + "learning_rate": 7.290143049351926e-06, + "loss": 0.0135, + "step": 82590 + }, + { + "epoch": 24.032586558044805, + "grad_norm": 0.1349404901266098, + "learning_rate": 7.286597493083255e-06, + "loss": 0.0148, + "step": 82600 + }, + { + "epoch": 24.03549607215595, + "grad_norm": 0.10768667608499527, + "learning_rate": 7.283052522596601e-06, + "loss": 0.0123, + "step": 82610 + }, + { + "epoch": 24.038405586267093, + "grad_norm": 0.09394870698451996, + "learning_rate": 7.279508138161188e-06, + "loss": 0.0135, + "step": 82620 + }, + { + "epoch": 24.041315100378238, + "grad_norm": 0.1280023753643036, + "learning_rate": 7.275964340046174e-06, + "loss": 0.014, + "step": 82630 + }, + { + "epoch": 24.04422461448938, + "grad_norm": 0.12544430792331696, + "learning_rate": 7.272421128520698e-06, + "loss": 0.0128, + "step": 82640 + }, + { + "epoch": 24.047134128600522, + "grad_norm": 0.11309689283370972, + "learning_rate": 7.268878503853835e-06, + "loss": 0.0125, + "step": 82650 + }, + { + "epoch": 24.050043642711668, + "grad_norm": 0.09666505455970764, + "learning_rate": 7.265336466314631e-06, + "loss": 0.0119, + "step": 82660 + }, + { + "epoch": 24.05295315682281, + "grad_norm": 0.09268681704998016, + "learning_rate": 7.261795016172082e-06, + "loss": 0.0125, + "step": 82670 + }, + { + "epoch": 24.055862670933955, + "grad_norm": 0.10470244288444519, + "learning_rate": 7.258254153695128e-06, + "loss": 0.0122, + "step": 82680 + }, + { + "epoch": 24.058772185045097, + "grad_norm": 0.10435837507247925, + "learning_rate": 7.2547138791526855e-06, + "loss": 0.0173, + "step": 82690 + }, + { + "epoch": 24.061681699156242, + "grad_norm": 0.14002597332000732, + "learning_rate": 7.2511741928136045e-06, + "loss": 0.014, + "step": 82700 + }, + { + "epoch": 24.064591213267384, + "grad_norm": 0.10739494860172272, + "learning_rate": 7.24763509494671e-06, + "loss": 0.013, + "step": 82710 + }, + { + "epoch": 24.067500727378526, + "grad_norm": 0.1249949187040329, + "learning_rate": 7.244096585820764e-06, + "loss": 0.0108, + "step": 82720 + }, + { + "epoch": 24.07041024148967, + "grad_norm": 0.13841913640499115, + "learning_rate": 7.240558665704498e-06, + "loss": 0.0115, + "step": 82730 + }, + { + "epoch": 24.073319755600814, + "grad_norm": 0.11475188285112381, + "learning_rate": 7.237021334866599e-06, + "loss": 0.0119, + "step": 82740 + }, + { + "epoch": 24.07622926971196, + "grad_norm": 0.12822413444519043, + "learning_rate": 7.2334845935757e-06, + "loss": 0.014, + "step": 82750 + }, + { + "epoch": 24.0791387838231, + "grad_norm": 0.11842960119247437, + "learning_rate": 7.229948442100379e-06, + "loss": 0.0129, + "step": 82760 + }, + { + "epoch": 24.082048297934247, + "grad_norm": 0.09881658852100372, + "learning_rate": 7.2264128807092055e-06, + "loss": 0.0115, + "step": 82770 + }, + { + "epoch": 24.08495781204539, + "grad_norm": 0.21406611800193787, + "learning_rate": 7.222877909670674e-06, + "loss": 0.0106, + "step": 82780 + }, + { + "epoch": 24.08786732615653, + "grad_norm": 0.10350848734378815, + "learning_rate": 7.219343529253235e-06, + "loss": 0.0127, + "step": 82790 + }, + { + "epoch": 24.090776840267676, + "grad_norm": 0.11177726835012436, + "learning_rate": 7.215809739725308e-06, + "loss": 0.0132, + "step": 82800 + }, + { + "epoch": 24.093686354378818, + "grad_norm": 0.1693701297044754, + "learning_rate": 7.212276541355266e-06, + "loss": 0.0161, + "step": 82810 + }, + { + "epoch": 24.096595868489963, + "grad_norm": 0.09501361846923828, + "learning_rate": 7.208743934411424e-06, + "loss": 0.0133, + "step": 82820 + }, + { + "epoch": 24.099505382601105, + "grad_norm": 0.11690454185009003, + "learning_rate": 7.205211919162058e-06, + "loss": 0.0115, + "step": 82830 + }, + { + "epoch": 24.10241489671225, + "grad_norm": 0.14387734234333038, + "learning_rate": 7.201680495875405e-06, + "loss": 0.0141, + "step": 82840 + }, + { + "epoch": 24.105324410823393, + "grad_norm": 0.12916497886180878, + "learning_rate": 7.1981496648196606e-06, + "loss": 0.0116, + "step": 82850 + }, + { + "epoch": 24.108233924934535, + "grad_norm": 0.165337935090065, + "learning_rate": 7.194619426262961e-06, + "loss": 0.0116, + "step": 82860 + }, + { + "epoch": 24.11114343904568, + "grad_norm": 0.10101848840713501, + "learning_rate": 7.191089780473393e-06, + "loss": 0.0122, + "step": 82870 + }, + { + "epoch": 24.114052953156822, + "grad_norm": 0.07546888291835785, + "learning_rate": 7.187560727719033e-06, + "loss": 0.0124, + "step": 82880 + }, + { + "epoch": 24.116962467267967, + "grad_norm": 0.1076006069779396, + "learning_rate": 7.184032268267879e-06, + "loss": 0.0147, + "step": 82890 + }, + { + "epoch": 24.11987198137911, + "grad_norm": 0.09824365377426147, + "learning_rate": 7.180504402387888e-06, + "loss": 0.0133, + "step": 82900 + }, + { + "epoch": 24.122781495490255, + "grad_norm": 0.10781807452440262, + "learning_rate": 7.176977130346983e-06, + "loss": 0.0135, + "step": 82910 + }, + { + "epoch": 24.125691009601397, + "grad_norm": 0.11603430658578873, + "learning_rate": 7.173450452413043e-06, + "loss": 0.0119, + "step": 82920 + }, + { + "epoch": 24.12860052371254, + "grad_norm": 0.09247943758964539, + "learning_rate": 7.169924368853884e-06, + "loss": 0.0122, + "step": 82930 + }, + { + "epoch": 24.131510037823684, + "grad_norm": 0.1023123636841774, + "learning_rate": 7.166398879937302e-06, + "loss": 0.012, + "step": 82940 + }, + { + "epoch": 24.134419551934826, + "grad_norm": 0.09903573989868164, + "learning_rate": 7.162873985931023e-06, + "loss": 0.0119, + "step": 82950 + }, + { + "epoch": 24.13732906604597, + "grad_norm": 0.15753623843193054, + "learning_rate": 7.159349687102748e-06, + "loss": 0.0124, + "step": 82960 + }, + { + "epoch": 24.140238580157114, + "grad_norm": 0.09723465144634247, + "learning_rate": 7.155825983720115e-06, + "loss": 0.0117, + "step": 82970 + }, + { + "epoch": 24.143148094268255, + "grad_norm": 0.09602390974760056, + "learning_rate": 7.152302876050733e-06, + "loss": 0.012, + "step": 82980 + }, + { + "epoch": 24.1460576083794, + "grad_norm": 0.12413492798805237, + "learning_rate": 7.148780364362162e-06, + "loss": 0.0141, + "step": 82990 + }, + { + "epoch": 24.148967122490543, + "grad_norm": 0.10741341859102249, + "learning_rate": 7.145258448921903e-06, + "loss": 0.0135, + "step": 83000 + }, + { + "epoch": 24.15187663660169, + "grad_norm": 0.11812126636505127, + "learning_rate": 7.141737129997433e-06, + "loss": 0.0119, + "step": 83010 + }, + { + "epoch": 24.15478615071283, + "grad_norm": 0.15825524926185608, + "learning_rate": 7.1382164078561636e-06, + "loss": 0.0122, + "step": 83020 + }, + { + "epoch": 24.157695664823976, + "grad_norm": 0.074445441365242, + "learning_rate": 7.134696282765473e-06, + "loss": 0.0109, + "step": 83030 + }, + { + "epoch": 24.160605178935118, + "grad_norm": 0.1313764899969101, + "learning_rate": 7.131176754992701e-06, + "loss": 0.0102, + "step": 83040 + }, + { + "epoch": 24.16351469304626, + "grad_norm": 0.12811067700386047, + "learning_rate": 7.1276578248051225e-06, + "loss": 0.0131, + "step": 83050 + }, + { + "epoch": 24.166424207157405, + "grad_norm": 0.09585830569267273, + "learning_rate": 7.124139492469977e-06, + "loss": 0.0115, + "step": 83060 + }, + { + "epoch": 24.169333721268547, + "grad_norm": 0.09484895318746567, + "learning_rate": 7.120621758254458e-06, + "loss": 0.0115, + "step": 83070 + }, + { + "epoch": 24.172243235379693, + "grad_norm": 0.10908060520887375, + "learning_rate": 7.117104622425724e-06, + "loss": 0.012, + "step": 83080 + }, + { + "epoch": 24.175152749490834, + "grad_norm": 0.08403359353542328, + "learning_rate": 7.113588085250865e-06, + "loss": 0.0134, + "step": 83090 + }, + { + "epoch": 24.17806226360198, + "grad_norm": 0.1226910948753357, + "learning_rate": 7.1100721469969445e-06, + "loss": 0.0111, + "step": 83100 + }, + { + "epoch": 24.180971777713122, + "grad_norm": 0.11065986752510071, + "learning_rate": 7.106556807930983e-06, + "loss": 0.0155, + "step": 83110 + }, + { + "epoch": 24.183881291824264, + "grad_norm": 0.08238428086042404, + "learning_rate": 7.103042068319938e-06, + "loss": 0.0112, + "step": 83120 + }, + { + "epoch": 24.18679080593541, + "grad_norm": 0.1462012678384781, + "learning_rate": 7.099527928430727e-06, + "loss": 0.0137, + "step": 83130 + }, + { + "epoch": 24.18970032004655, + "grad_norm": 0.09594299644231796, + "learning_rate": 7.096014388530232e-06, + "loss": 0.0108, + "step": 83140 + }, + { + "epoch": 24.192609834157697, + "grad_norm": 0.10113846510648727, + "learning_rate": 7.092501448885289e-06, + "loss": 0.0116, + "step": 83150 + }, + { + "epoch": 24.19551934826884, + "grad_norm": 0.09511218965053558, + "learning_rate": 7.088989109762675e-06, + "loss": 0.0118, + "step": 83160 + }, + { + "epoch": 24.198428862379984, + "grad_norm": 0.11964470148086548, + "learning_rate": 7.085477371429119e-06, + "loss": 0.0133, + "step": 83170 + }, + { + "epoch": 24.201338376491126, + "grad_norm": 0.1072021871805191, + "learning_rate": 7.081966234151336e-06, + "loss": 0.0133, + "step": 83180 + }, + { + "epoch": 24.204247890602268, + "grad_norm": 0.07359758764505386, + "learning_rate": 7.078455698195963e-06, + "loss": 0.0117, + "step": 83190 + }, + { + "epoch": 24.207157404713413, + "grad_norm": 0.08812130987644196, + "learning_rate": 7.0749457638295975e-06, + "loss": 0.0127, + "step": 83200 + }, + { + "epoch": 24.210066918824555, + "grad_norm": 0.09301616996526718, + "learning_rate": 7.071436431318798e-06, + "loss": 0.011, + "step": 83210 + }, + { + "epoch": 24.2129764329357, + "grad_norm": 0.12970015406608582, + "learning_rate": 7.067927700930085e-06, + "loss": 0.0115, + "step": 83220 + }, + { + "epoch": 24.215885947046843, + "grad_norm": 0.17245851457118988, + "learning_rate": 7.064419572929911e-06, + "loss": 0.0112, + "step": 83230 + }, + { + "epoch": 24.21879546115799, + "grad_norm": 0.07814918458461761, + "learning_rate": 7.060912047584704e-06, + "loss": 0.0125, + "step": 83240 + }, + { + "epoch": 24.22170497526913, + "grad_norm": 0.1193484514951706, + "learning_rate": 7.057405125160829e-06, + "loss": 0.0141, + "step": 83250 + }, + { + "epoch": 24.224614489380272, + "grad_norm": 0.08192596584558487, + "learning_rate": 7.053898805924625e-06, + "loss": 0.0107, + "step": 83260 + }, + { + "epoch": 24.227524003491418, + "grad_norm": 0.14677855372428894, + "learning_rate": 7.0503930901423596e-06, + "loss": 0.0116, + "step": 83270 + }, + { + "epoch": 24.23043351760256, + "grad_norm": 0.1094379872083664, + "learning_rate": 7.046887978080276e-06, + "loss": 0.0116, + "step": 83280 + }, + { + "epoch": 24.233343031713705, + "grad_norm": 0.17159543931484222, + "learning_rate": 7.04338347000457e-06, + "loss": 0.0132, + "step": 83290 + }, + { + "epoch": 24.236252545824847, + "grad_norm": 0.10508960485458374, + "learning_rate": 7.039879566181377e-06, + "loss": 0.0134, + "step": 83300 + }, + { + "epoch": 24.23916205993599, + "grad_norm": 0.09240979701280594, + "learning_rate": 7.036376266876802e-06, + "loss": 0.0141, + "step": 83310 + }, + { + "epoch": 24.242071574047134, + "grad_norm": 0.13323789834976196, + "learning_rate": 7.03287357235689e-06, + "loss": 0.0133, + "step": 83320 + }, + { + "epoch": 24.244981088158276, + "grad_norm": 0.09417569637298584, + "learning_rate": 7.029371482887656e-06, + "loss": 0.0122, + "step": 83330 + }, + { + "epoch": 24.247890602269422, + "grad_norm": 0.11300661414861679, + "learning_rate": 7.025869998735052e-06, + "loss": 0.0118, + "step": 83340 + }, + { + "epoch": 24.250800116380564, + "grad_norm": 0.09162361919879913, + "learning_rate": 7.022369120165003e-06, + "loss": 0.0128, + "step": 83350 + }, + { + "epoch": 24.25370963049171, + "grad_norm": 0.09871980547904968, + "learning_rate": 7.018868847443365e-06, + "loss": 0.0127, + "step": 83360 + }, + { + "epoch": 24.25661914460285, + "grad_norm": 0.15046939253807068, + "learning_rate": 7.0153691808359675e-06, + "loss": 0.0109, + "step": 83370 + }, + { + "epoch": 24.259528658713993, + "grad_norm": 0.10397671908140182, + "learning_rate": 7.01187012060859e-06, + "loss": 0.0132, + "step": 83380 + }, + { + "epoch": 24.26243817282514, + "grad_norm": 0.10557131469249725, + "learning_rate": 7.008371667026958e-06, + "loss": 0.012, + "step": 83390 + }, + { + "epoch": 24.26534768693628, + "grad_norm": 0.10859280824661255, + "learning_rate": 7.004873820356754e-06, + "loss": 0.013, + "step": 83400 + }, + { + "epoch": 24.268257201047426, + "grad_norm": 0.10231134295463562, + "learning_rate": 7.0013765808636295e-06, + "loss": 0.0123, + "step": 83410 + }, + { + "epoch": 24.271166715158568, + "grad_norm": 0.08701319992542267, + "learning_rate": 6.9978799488131635e-06, + "loss": 0.0115, + "step": 83420 + }, + { + "epoch": 24.274076229269713, + "grad_norm": 0.10420706123113632, + "learning_rate": 6.994383924470904e-06, + "loss": 0.0128, + "step": 83430 + }, + { + "epoch": 24.276985743380855, + "grad_norm": 0.08345292508602142, + "learning_rate": 6.990888508102351e-06, + "loss": 0.0106, + "step": 83440 + }, + { + "epoch": 24.279895257491997, + "grad_norm": 0.1129554957151413, + "learning_rate": 6.987393699972966e-06, + "loss": 0.0125, + "step": 83450 + }, + { + "epoch": 24.282804771603143, + "grad_norm": 0.13869084417819977, + "learning_rate": 6.98389950034815e-06, + "loss": 0.0109, + "step": 83460 + }, + { + "epoch": 24.285714285714285, + "grad_norm": 0.10070401430130005, + "learning_rate": 6.980405909493255e-06, + "loss": 0.0119, + "step": 83470 + }, + { + "epoch": 24.28862379982543, + "grad_norm": 0.08557243645191193, + "learning_rate": 6.976912927673619e-06, + "loss": 0.0111, + "step": 83480 + }, + { + "epoch": 24.291533313936572, + "grad_norm": 0.15485282242298126, + "learning_rate": 6.973420555154496e-06, + "loss": 0.0104, + "step": 83490 + }, + { + "epoch": 24.294442828047718, + "grad_norm": 0.21515241265296936, + "learning_rate": 6.969928792201108e-06, + "loss": 0.0129, + "step": 83500 + }, + { + "epoch": 24.29735234215886, + "grad_norm": 0.11799183487892151, + "learning_rate": 6.966437639078632e-06, + "loss": 0.0141, + "step": 83510 + }, + { + "epoch": 24.30026185627, + "grad_norm": 0.07658097892999649, + "learning_rate": 6.9629470960522065e-06, + "loss": 0.0111, + "step": 83520 + }, + { + "epoch": 24.303171370381147, + "grad_norm": 0.08752334117889404, + "learning_rate": 6.95945716338691e-06, + "loss": 0.012, + "step": 83530 + }, + { + "epoch": 24.30608088449229, + "grad_norm": 0.10720223933458328, + "learning_rate": 6.955967841347772e-06, + "loss": 0.0126, + "step": 83540 + }, + { + "epoch": 24.308990398603434, + "grad_norm": 0.12938199937343597, + "learning_rate": 6.9524791301997924e-06, + "loss": 0.011, + "step": 83550 + }, + { + "epoch": 24.311899912714576, + "grad_norm": 0.11460397392511368, + "learning_rate": 6.948991030207918e-06, + "loss": 0.013, + "step": 83560 + }, + { + "epoch": 24.31480942682572, + "grad_norm": 0.09389664232730865, + "learning_rate": 6.945503541637038e-06, + "loss": 0.0116, + "step": 83570 + }, + { + "epoch": 24.317718940936864, + "grad_norm": 0.1028117761015892, + "learning_rate": 6.942016664752009e-06, + "loss": 0.0101, + "step": 83580 + }, + { + "epoch": 24.320628455048006, + "grad_norm": 0.1144455298781395, + "learning_rate": 6.938530399817642e-06, + "loss": 0.0127, + "step": 83590 + }, + { + "epoch": 24.32353796915915, + "grad_norm": 0.09642479568719864, + "learning_rate": 6.935044747098684e-06, + "loss": 0.013, + "step": 83600 + }, + { + "epoch": 24.326447483270293, + "grad_norm": 0.18028739094734192, + "learning_rate": 6.931559706859857e-06, + "loss": 0.0106, + "step": 83610 + }, + { + "epoch": 24.32935699738144, + "grad_norm": 0.10355337709188461, + "learning_rate": 6.928075279365821e-06, + "loss": 0.012, + "step": 83620 + }, + { + "epoch": 24.33226651149258, + "grad_norm": 0.10301706939935684, + "learning_rate": 6.924591464881201e-06, + "loss": 0.0139, + "step": 83630 + }, + { + "epoch": 24.335176025603722, + "grad_norm": 0.0841703861951828, + "learning_rate": 6.921108263670561e-06, + "loss": 0.011, + "step": 83640 + }, + { + "epoch": 24.338085539714868, + "grad_norm": 0.11399044841527939, + "learning_rate": 6.917625675998439e-06, + "loss": 0.0112, + "step": 83650 + }, + { + "epoch": 24.34099505382601, + "grad_norm": 0.09500688314437866, + "learning_rate": 6.914143702129301e-06, + "loss": 0.0119, + "step": 83660 + }, + { + "epoch": 24.343904567937155, + "grad_norm": 0.10691593587398529, + "learning_rate": 6.910662342327588e-06, + "loss": 0.0137, + "step": 83670 + }, + { + "epoch": 24.346814082048297, + "grad_norm": 0.16883257031440735, + "learning_rate": 6.907181596857688e-06, + "loss": 0.0129, + "step": 83680 + }, + { + "epoch": 24.349723596159443, + "grad_norm": 0.14022555947303772, + "learning_rate": 6.903701465983935e-06, + "loss": 0.0126, + "step": 83690 + }, + { + "epoch": 24.352633110270585, + "grad_norm": 0.11728333681821823, + "learning_rate": 6.900221949970629e-06, + "loss": 0.0114, + "step": 83700 + }, + { + "epoch": 24.355542624381727, + "grad_norm": 0.11998418718576431, + "learning_rate": 6.896743049082007e-06, + "loss": 0.0127, + "step": 83710 + }, + { + "epoch": 24.358452138492872, + "grad_norm": 0.11773742735385895, + "learning_rate": 6.893264763582279e-06, + "loss": 0.0136, + "step": 83720 + }, + { + "epoch": 24.361361652604014, + "grad_norm": 0.16111917793750763, + "learning_rate": 6.889787093735587e-06, + "loss": 0.0119, + "step": 83730 + }, + { + "epoch": 24.36427116671516, + "grad_norm": 0.08807458728551865, + "learning_rate": 6.886310039806043e-06, + "loss": 0.0104, + "step": 83740 + }, + { + "epoch": 24.3671806808263, + "grad_norm": 0.0880085676908493, + "learning_rate": 6.88283360205771e-06, + "loss": 0.0114, + "step": 83750 + }, + { + "epoch": 24.370090194937447, + "grad_norm": 0.09356711804866791, + "learning_rate": 6.8793577807545966e-06, + "loss": 0.0116, + "step": 83760 + }, + { + "epoch": 24.37299970904859, + "grad_norm": 0.1082274317741394, + "learning_rate": 6.875882576160657e-06, + "loss": 0.0124, + "step": 83770 + }, + { + "epoch": 24.37590922315973, + "grad_norm": 0.0960974246263504, + "learning_rate": 6.872407988539833e-06, + "loss": 0.0131, + "step": 83780 + }, + { + "epoch": 24.378818737270876, + "grad_norm": 0.11708246171474457, + "learning_rate": 6.8689340181559844e-06, + "loss": 0.013, + "step": 83790 + }, + { + "epoch": 24.381728251382018, + "grad_norm": 0.11301849037408829, + "learning_rate": 6.8654606652729315e-06, + "loss": 0.0116, + "step": 83800 + }, + { + "epoch": 24.384637765493164, + "grad_norm": 0.09922720491886139, + "learning_rate": 6.861987930154457e-06, + "loss": 0.0106, + "step": 83810 + }, + { + "epoch": 24.387547279604306, + "grad_norm": 0.12191013246774673, + "learning_rate": 6.8585158130643e-06, + "loss": 0.011, + "step": 83820 + }, + { + "epoch": 24.39045679371545, + "grad_norm": 0.12471474707126617, + "learning_rate": 6.855044314266138e-06, + "loss": 0.0123, + "step": 83830 + }, + { + "epoch": 24.393366307826593, + "grad_norm": 0.13783583045005798, + "learning_rate": 6.851573434023602e-06, + "loss": 0.0105, + "step": 83840 + }, + { + "epoch": 24.396275821937735, + "grad_norm": 0.12884919345378876, + "learning_rate": 6.848103172600287e-06, + "loss": 0.0127, + "step": 83850 + }, + { + "epoch": 24.39918533604888, + "grad_norm": 0.08345529437065125, + "learning_rate": 6.844633530259747e-06, + "loss": 0.0125, + "step": 83860 + }, + { + "epoch": 24.402094850160022, + "grad_norm": 0.09292532503604889, + "learning_rate": 6.841164507265461e-06, + "loss": 0.0116, + "step": 83870 + }, + { + "epoch": 24.405004364271168, + "grad_norm": 0.0853348895907402, + "learning_rate": 6.8376961038808886e-06, + "loss": 0.0132, + "step": 83880 + }, + { + "epoch": 24.40791387838231, + "grad_norm": 0.09345197677612305, + "learning_rate": 6.834228320369436e-06, + "loss": 0.0132, + "step": 83890 + }, + { + "epoch": 24.410823392493455, + "grad_norm": 0.10964822769165039, + "learning_rate": 6.830761156994454e-06, + "loss": 0.0123, + "step": 83900 + }, + { + "epoch": 24.413732906604597, + "grad_norm": 0.07953955233097076, + "learning_rate": 6.827294614019243e-06, + "loss": 0.0112, + "step": 83910 + }, + { + "epoch": 24.41664242071574, + "grad_norm": 0.09500192105770111, + "learning_rate": 6.82382869170707e-06, + "loss": 0.0109, + "step": 83920 + }, + { + "epoch": 24.419551934826885, + "grad_norm": 0.13088753819465637, + "learning_rate": 6.820363390321158e-06, + "loss": 0.0117, + "step": 83930 + }, + { + "epoch": 24.422461448938027, + "grad_norm": 0.10180121660232544, + "learning_rate": 6.816898710124658e-06, + "loss": 0.0124, + "step": 83940 + }, + { + "epoch": 24.425370963049172, + "grad_norm": 0.1419447809457779, + "learning_rate": 6.8134346513807035e-06, + "loss": 0.0121, + "step": 83950 + }, + { + "epoch": 24.428280477160314, + "grad_norm": 0.13770371675491333, + "learning_rate": 6.8099712143523555e-06, + "loss": 0.0135, + "step": 83960 + }, + { + "epoch": 24.43118999127146, + "grad_norm": 0.07930774986743927, + "learning_rate": 6.806508399302644e-06, + "loss": 0.0113, + "step": 83970 + }, + { + "epoch": 24.4340995053826, + "grad_norm": 0.10922575742006302, + "learning_rate": 6.803046206494553e-06, + "loss": 0.0142, + "step": 83980 + }, + { + "epoch": 24.437009019493743, + "grad_norm": 0.10654281824827194, + "learning_rate": 6.7995846361910015e-06, + "loss": 0.0134, + "step": 83990 + }, + { + "epoch": 24.43991853360489, + "grad_norm": 0.12408339232206345, + "learning_rate": 6.796123688654883e-06, + "loss": 0.0115, + "step": 84000 + }, + { + "epoch": 24.44282804771603, + "grad_norm": 0.08873133361339569, + "learning_rate": 6.792663364149025e-06, + "loss": 0.0109, + "step": 84010 + }, + { + "epoch": 24.445737561827176, + "grad_norm": 0.092460498213768, + "learning_rate": 6.789203662936227e-06, + "loss": 0.0125, + "step": 84020 + }, + { + "epoch": 24.448647075938318, + "grad_norm": 0.13622546195983887, + "learning_rate": 6.785744585279219e-06, + "loss": 0.0137, + "step": 84030 + }, + { + "epoch": 24.45155659004946, + "grad_norm": 0.12668123841285706, + "learning_rate": 6.7822861314406986e-06, + "loss": 0.0118, + "step": 84040 + }, + { + "epoch": 24.454466104160606, + "grad_norm": 0.08967998623847961, + "learning_rate": 6.778828301683322e-06, + "loss": 0.0129, + "step": 84050 + }, + { + "epoch": 24.457375618271747, + "grad_norm": 0.14119884371757507, + "learning_rate": 6.775371096269681e-06, + "loss": 0.0162, + "step": 84060 + }, + { + "epoch": 24.460285132382893, + "grad_norm": 0.10737556964159012, + "learning_rate": 6.771914515462319e-06, + "loss": 0.0106, + "step": 84070 + }, + { + "epoch": 24.463194646494035, + "grad_norm": 0.09337040036916733, + "learning_rate": 6.76845855952375e-06, + "loss": 0.0107, + "step": 84080 + }, + { + "epoch": 24.46610416060518, + "grad_norm": 0.10821323841810226, + "learning_rate": 6.765003228716437e-06, + "loss": 0.012, + "step": 84090 + }, + { + "epoch": 24.469013674716322, + "grad_norm": 0.1224534884095192, + "learning_rate": 6.761548523302776e-06, + "loss": 0.012, + "step": 84100 + }, + { + "epoch": 24.471923188827464, + "grad_norm": 0.12908561527729034, + "learning_rate": 6.758094443545136e-06, + "loss": 0.0109, + "step": 84110 + }, + { + "epoch": 24.47483270293861, + "grad_norm": 0.1298784762620926, + "learning_rate": 6.754640989705836e-06, + "loss": 0.0135, + "step": 84120 + }, + { + "epoch": 24.47774221704975, + "grad_norm": 0.10730090737342834, + "learning_rate": 6.75118816204714e-06, + "loss": 0.0124, + "step": 84130 + }, + { + "epoch": 24.480651731160897, + "grad_norm": 0.06932874768972397, + "learning_rate": 6.747735960831258e-06, + "loss": 0.011, + "step": 84140 + }, + { + "epoch": 24.48356124527204, + "grad_norm": 0.10484591126441956, + "learning_rate": 6.74428438632037e-06, + "loss": 0.0127, + "step": 84150 + }, + { + "epoch": 24.486470759383185, + "grad_norm": 0.0943303033709526, + "learning_rate": 6.740833438776605e-06, + "loss": 0.0119, + "step": 84160 + }, + { + "epoch": 24.489380273494326, + "grad_norm": 0.1952827125787735, + "learning_rate": 6.737383118462036e-06, + "loss": 0.0112, + "step": 84170 + }, + { + "epoch": 24.49228978760547, + "grad_norm": 0.1292068064212799, + "learning_rate": 6.733933425638679e-06, + "loss": 0.013, + "step": 84180 + }, + { + "epoch": 24.495199301716614, + "grad_norm": 0.09891398996114731, + "learning_rate": 6.730484360568537e-06, + "loss": 0.0152, + "step": 84190 + }, + { + "epoch": 24.498108815827756, + "grad_norm": 0.10014667361974716, + "learning_rate": 6.727035923513536e-06, + "loss": 0.0131, + "step": 84200 + }, + { + "epoch": 24.5010183299389, + "grad_norm": 0.09826000779867172, + "learning_rate": 6.723588114735552e-06, + "loss": 0.0118, + "step": 84210 + }, + { + "epoch": 24.503927844050043, + "grad_norm": 0.12038980424404144, + "learning_rate": 6.720140934496432e-06, + "loss": 0.0125, + "step": 84220 + }, + { + "epoch": 24.50683735816119, + "grad_norm": 0.09138856828212738, + "learning_rate": 6.716694383057971e-06, + "loss": 0.0114, + "step": 84230 + }, + { + "epoch": 24.50974687227233, + "grad_norm": 0.18325775861740112, + "learning_rate": 6.713248460681903e-06, + "loss": 0.0132, + "step": 84240 + }, + { + "epoch": 24.512656386383473, + "grad_norm": 0.16167372465133667, + "learning_rate": 6.709803167629929e-06, + "loss": 0.0122, + "step": 84250 + }, + { + "epoch": 24.515565900494618, + "grad_norm": 0.12649112939834595, + "learning_rate": 6.70635850416369e-06, + "loss": 0.0142, + "step": 84260 + }, + { + "epoch": 24.51847541460576, + "grad_norm": 0.11901970952749252, + "learning_rate": 6.702914470544795e-06, + "loss": 0.0111, + "step": 84270 + }, + { + "epoch": 24.521384928716905, + "grad_norm": 0.11200658231973648, + "learning_rate": 6.699471067034784e-06, + "loss": 0.0106, + "step": 84280 + }, + { + "epoch": 24.524294442828047, + "grad_norm": 0.11619196087121964, + "learning_rate": 6.696028293895167e-06, + "loss": 0.0102, + "step": 84290 + }, + { + "epoch": 24.527203956939193, + "grad_norm": 0.08826770633459091, + "learning_rate": 6.6925861513874044e-06, + "loss": 0.0118, + "step": 84300 + }, + { + "epoch": 24.530113471050335, + "grad_norm": 0.12257298082113266, + "learning_rate": 6.689144639772895e-06, + "loss": 0.0128, + "step": 84310 + }, + { + "epoch": 24.533022985161477, + "grad_norm": 0.12080123275518417, + "learning_rate": 6.685703759313009e-06, + "loss": 0.0132, + "step": 84320 + }, + { + "epoch": 24.535932499272622, + "grad_norm": 0.0904434472322464, + "learning_rate": 6.682263510269045e-06, + "loss": 0.0109, + "step": 84330 + }, + { + "epoch": 24.538842013383764, + "grad_norm": 0.10573882609605789, + "learning_rate": 6.6788238929022775e-06, + "loss": 0.0119, + "step": 84340 + }, + { + "epoch": 24.54175152749491, + "grad_norm": 0.11669043451547623, + "learning_rate": 6.675384907473927e-06, + "loss": 0.0102, + "step": 84350 + }, + { + "epoch": 24.54466104160605, + "grad_norm": 0.1005919948220253, + "learning_rate": 6.671946554245153e-06, + "loss": 0.0125, + "step": 84360 + }, + { + "epoch": 24.547570555717193, + "grad_norm": 0.11740623414516449, + "learning_rate": 6.668508833477072e-06, + "loss": 0.013, + "step": 84370 + }, + { + "epoch": 24.55048006982834, + "grad_norm": 0.12094961106777191, + "learning_rate": 6.665071745430762e-06, + "loss": 0.0107, + "step": 84380 + }, + { + "epoch": 24.55338958393948, + "grad_norm": 0.06900829076766968, + "learning_rate": 6.661635290367254e-06, + "loss": 0.0106, + "step": 84390 + }, + { + "epoch": 24.556299098050626, + "grad_norm": 0.11806595325469971, + "learning_rate": 6.6581994685475105e-06, + "loss": 0.0107, + "step": 84400 + }, + { + "epoch": 24.55920861216177, + "grad_norm": 0.09764142334461212, + "learning_rate": 6.6547642802324684e-06, + "loss": 0.0103, + "step": 84410 + }, + { + "epoch": 24.562118126272914, + "grad_norm": 0.0978640541434288, + "learning_rate": 6.651329725683009e-06, + "loss": 0.0104, + "step": 84420 + }, + { + "epoch": 24.565027640384056, + "grad_norm": 0.08588864654302597, + "learning_rate": 6.647895805159962e-06, + "loss": 0.0103, + "step": 84430 + }, + { + "epoch": 24.567937154495198, + "grad_norm": 0.2674052119255066, + "learning_rate": 6.644462518924096e-06, + "loss": 0.0138, + "step": 84440 + }, + { + "epoch": 24.570846668606343, + "grad_norm": 0.10221811383962631, + "learning_rate": 6.641029867236173e-06, + "loss": 0.0132, + "step": 84450 + }, + { + "epoch": 24.573756182717485, + "grad_norm": 0.1357974261045456, + "learning_rate": 6.637597850356867e-06, + "loss": 0.0122, + "step": 84460 + }, + { + "epoch": 24.57666569682863, + "grad_norm": 0.098868228495121, + "learning_rate": 6.634166468546816e-06, + "loss": 0.0108, + "step": 84470 + }, + { + "epoch": 24.579575210939772, + "grad_norm": 0.05891802906990051, + "learning_rate": 6.630735722066602e-06, + "loss": 0.013, + "step": 84480 + }, + { + "epoch": 24.582484725050918, + "grad_norm": 0.21239420771598816, + "learning_rate": 6.627305611176788e-06, + "loss": 0.0114, + "step": 84490 + }, + { + "epoch": 24.58539423916206, + "grad_norm": 0.09122376143932343, + "learning_rate": 6.62387613613786e-06, + "loss": 0.0121, + "step": 84500 + }, + { + "epoch": 24.588303753273202, + "grad_norm": 0.09854885190725327, + "learning_rate": 6.620447297210253e-06, + "loss": 0.0121, + "step": 84510 + }, + { + "epoch": 24.591213267384347, + "grad_norm": 0.17235167324543, + "learning_rate": 6.6170190946543765e-06, + "loss": 0.0118, + "step": 84520 + }, + { + "epoch": 24.59412278149549, + "grad_norm": 0.16576963663101196, + "learning_rate": 6.613591528730581e-06, + "loss": 0.0132, + "step": 84530 + }, + { + "epoch": 24.597032295606635, + "grad_norm": 0.08923134952783585, + "learning_rate": 6.610164599699164e-06, + "loss": 0.0128, + "step": 84540 + }, + { + "epoch": 24.599941809717777, + "grad_norm": 0.08264674246311188, + "learning_rate": 6.606738307820366e-06, + "loss": 0.0104, + "step": 84550 + }, + { + "epoch": 24.602851323828922, + "grad_norm": 0.09061755239963531, + "learning_rate": 6.603312653354414e-06, + "loss": 0.0108, + "step": 84560 + }, + { + "epoch": 24.605760837940064, + "grad_norm": 0.09853769838809967, + "learning_rate": 6.599887636561452e-06, + "loss": 0.011, + "step": 84570 + }, + { + "epoch": 24.608670352051206, + "grad_norm": 0.09948179870843887, + "learning_rate": 6.5964632577015824e-06, + "loss": 0.0125, + "step": 84580 + }, + { + "epoch": 24.61157986616235, + "grad_norm": 0.16983796656131744, + "learning_rate": 6.593039517034871e-06, + "loss": 0.0125, + "step": 84590 + }, + { + "epoch": 24.614489380273493, + "grad_norm": 0.19061626493930817, + "learning_rate": 6.5896164148213314e-06, + "loss": 0.0112, + "step": 84600 + }, + { + "epoch": 24.61739889438464, + "grad_norm": 0.10366039723157883, + "learning_rate": 6.586193951320918e-06, + "loss": 0.0117, + "step": 84610 + }, + { + "epoch": 24.62030840849578, + "grad_norm": 0.15226268768310547, + "learning_rate": 6.582772126793552e-06, + "loss": 0.012, + "step": 84620 + }, + { + "epoch": 24.623217922606926, + "grad_norm": 0.08066537976264954, + "learning_rate": 6.579350941499088e-06, + "loss": 0.0125, + "step": 84630 + }, + { + "epoch": 24.62612743671807, + "grad_norm": 0.10176479816436768, + "learning_rate": 6.575930395697356e-06, + "loss": 0.0125, + "step": 84640 + }, + { + "epoch": 24.62903695082921, + "grad_norm": 0.09698158502578735, + "learning_rate": 6.572510489648108e-06, + "loss": 0.0141, + "step": 84650 + }, + { + "epoch": 24.631946464940356, + "grad_norm": 0.13663044571876526, + "learning_rate": 6.569091223611073e-06, + "loss": 0.0137, + "step": 84660 + }, + { + "epoch": 24.634855979051498, + "grad_norm": 0.10159241408109665, + "learning_rate": 6.565672597845927e-06, + "loss": 0.0126, + "step": 84670 + }, + { + "epoch": 24.637765493162643, + "grad_norm": 0.07903256267309189, + "learning_rate": 6.562254612612279e-06, + "loss": 0.0117, + "step": 84680 + }, + { + "epoch": 24.640675007273785, + "grad_norm": 0.09893683344125748, + "learning_rate": 6.558837268169715e-06, + "loss": 0.0125, + "step": 84690 + }, + { + "epoch": 24.64358452138493, + "grad_norm": 0.11926772445440292, + "learning_rate": 6.555420564777748e-06, + "loss": 0.0138, + "step": 84700 + }, + { + "epoch": 24.646494035496072, + "grad_norm": 0.10656420886516571, + "learning_rate": 6.552004502695859e-06, + "loss": 0.0119, + "step": 84710 + }, + { + "epoch": 24.649403549607214, + "grad_norm": 0.13725194334983826, + "learning_rate": 6.548589082183481e-06, + "loss": 0.0124, + "step": 84720 + }, + { + "epoch": 24.65231306371836, + "grad_norm": 0.09096939116716385, + "learning_rate": 6.545174303499989e-06, + "loss": 0.014, + "step": 84730 + }, + { + "epoch": 24.6552225778295, + "grad_norm": 0.07673488557338715, + "learning_rate": 6.541760166904705e-06, + "loss": 0.0114, + "step": 84740 + }, + { + "epoch": 24.658132091940647, + "grad_norm": 0.08832289278507233, + "learning_rate": 6.538346672656919e-06, + "loss": 0.0116, + "step": 84750 + }, + { + "epoch": 24.66104160605179, + "grad_norm": 0.14008720219135284, + "learning_rate": 6.5349338210158654e-06, + "loss": 0.0122, + "step": 84760 + }, + { + "epoch": 24.66395112016293, + "grad_norm": 0.10672137886285782, + "learning_rate": 6.5315216122407176e-06, + "loss": 0.0112, + "step": 84770 + }, + { + "epoch": 24.666860634274077, + "grad_norm": 0.1258135437965393, + "learning_rate": 6.528110046590617e-06, + "loss": 0.0136, + "step": 84780 + }, + { + "epoch": 24.66977014838522, + "grad_norm": 0.12626062333583832, + "learning_rate": 6.5246991243246535e-06, + "loss": 0.0127, + "step": 84790 + }, + { + "epoch": 24.672679662496364, + "grad_norm": 0.0899471864104271, + "learning_rate": 6.52128884570186e-06, + "loss": 0.0107, + "step": 84800 + }, + { + "epoch": 24.675589176607506, + "grad_norm": 0.08660640567541122, + "learning_rate": 6.517879210981218e-06, + "loss": 0.0129, + "step": 84810 + }, + { + "epoch": 24.67849869071865, + "grad_norm": 0.14390166103839874, + "learning_rate": 6.514470220421674e-06, + "loss": 0.0124, + "step": 84820 + }, + { + "epoch": 24.681408204829793, + "grad_norm": 0.0861000046133995, + "learning_rate": 6.511061874282122e-06, + "loss": 0.0146, + "step": 84830 + }, + { + "epoch": 24.684317718940935, + "grad_norm": 0.08985184133052826, + "learning_rate": 6.507654172821397e-06, + "loss": 0.0119, + "step": 84840 + }, + { + "epoch": 24.68722723305208, + "grad_norm": 0.10922583192586899, + "learning_rate": 6.504247116298284e-06, + "loss": 0.0127, + "step": 84850 + }, + { + "epoch": 24.690136747163223, + "grad_norm": 0.08032093942165375, + "learning_rate": 6.500840704971546e-06, + "loss": 0.0109, + "step": 84860 + }, + { + "epoch": 24.693046261274368, + "grad_norm": 0.1039203554391861, + "learning_rate": 6.4974349390998665e-06, + "loss": 0.0119, + "step": 84870 + }, + { + "epoch": 24.69595577538551, + "grad_norm": 0.1324070692062378, + "learning_rate": 6.494029818941885e-06, + "loss": 0.0115, + "step": 84880 + }, + { + "epoch": 24.698865289496656, + "grad_norm": 0.08818351477384567, + "learning_rate": 6.490625344756206e-06, + "loss": 0.0111, + "step": 84890 + }, + { + "epoch": 24.701774803607798, + "grad_norm": 0.1277056336402893, + "learning_rate": 6.4872215168013804e-06, + "loss": 0.0119, + "step": 84900 + }, + { + "epoch": 24.70468431771894, + "grad_norm": 0.12060955911874771, + "learning_rate": 6.4838183353359e-06, + "loss": 0.0131, + "step": 84910 + }, + { + "epoch": 24.707593831830085, + "grad_norm": 0.09583266824483871, + "learning_rate": 6.480415800618211e-06, + "loss": 0.011, + "step": 84920 + }, + { + "epoch": 24.710503345941227, + "grad_norm": 0.11352371424436569, + "learning_rate": 6.477013912906716e-06, + "loss": 0.0113, + "step": 84930 + }, + { + "epoch": 24.713412860052372, + "grad_norm": 0.1432604044675827, + "learning_rate": 6.473612672459774e-06, + "loss": 0.0118, + "step": 84940 + }, + { + "epoch": 24.716322374163514, + "grad_norm": 0.1313677281141281, + "learning_rate": 6.470212079535674e-06, + "loss": 0.0127, + "step": 84950 + }, + { + "epoch": 24.71923188827466, + "grad_norm": 0.09042922407388687, + "learning_rate": 6.466812134392675e-06, + "loss": 0.011, + "step": 84960 + }, + { + "epoch": 24.7221414023858, + "grad_norm": 0.12830260396003723, + "learning_rate": 6.463412837288985e-06, + "loss": 0.0118, + "step": 84970 + }, + { + "epoch": 24.725050916496944, + "grad_norm": 0.1021689772605896, + "learning_rate": 6.4600141884827505e-06, + "loss": 0.0111, + "step": 84980 + }, + { + "epoch": 24.72796043060809, + "grad_norm": 0.1302712857723236, + "learning_rate": 6.456616188232082e-06, + "loss": 0.0117, + "step": 84990 + }, + { + "epoch": 24.73086994471923, + "grad_norm": 0.078711599111557, + "learning_rate": 6.453218836795029e-06, + "loss": 0.0123, + "step": 85000 + }, + { + "epoch": 24.733779458830377, + "grad_norm": 0.07898242771625519, + "learning_rate": 6.449822134429607e-06, + "loss": 0.012, + "step": 85010 + }, + { + "epoch": 24.73668897294152, + "grad_norm": 0.14702893793582916, + "learning_rate": 6.446426081393762e-06, + "loss": 0.0124, + "step": 85020 + }, + { + "epoch": 24.73959848705266, + "grad_norm": 0.16647924482822418, + "learning_rate": 6.443030677945413e-06, + "loss": 0.0158, + "step": 85030 + }, + { + "epoch": 24.742508001163806, + "grad_norm": 0.11932359635829926, + "learning_rate": 6.439635924342405e-06, + "loss": 0.0114, + "step": 85040 + }, + { + "epoch": 24.745417515274948, + "grad_norm": 0.10614825040102005, + "learning_rate": 6.436241820842558e-06, + "loss": 0.0128, + "step": 85050 + }, + { + "epoch": 24.748327029386093, + "grad_norm": 0.1117333173751831, + "learning_rate": 6.432848367703634e-06, + "loss": 0.012, + "step": 85060 + }, + { + "epoch": 24.751236543497235, + "grad_norm": 0.11020760983228683, + "learning_rate": 6.429455565183332e-06, + "loss": 0.013, + "step": 85070 + }, + { + "epoch": 24.75414605760838, + "grad_norm": 0.2211574912071228, + "learning_rate": 6.426063413539318e-06, + "loss": 0.0114, + "step": 85080 + }, + { + "epoch": 24.757055571719523, + "grad_norm": 0.1182866096496582, + "learning_rate": 6.422671913029213e-06, + "loss": 0.011, + "step": 85090 + }, + { + "epoch": 24.759965085830665, + "grad_norm": 0.1298506259918213, + "learning_rate": 6.419281063910568e-06, + "loss": 0.0112, + "step": 85100 + }, + { + "epoch": 24.76287459994181, + "grad_norm": 0.10450176894664764, + "learning_rate": 6.415890866440893e-06, + "loss": 0.0124, + "step": 85110 + }, + { + "epoch": 24.765784114052952, + "grad_norm": 0.13375215232372284, + "learning_rate": 6.412501320877659e-06, + "loss": 0.0106, + "step": 85120 + }, + { + "epoch": 24.768693628164097, + "grad_norm": 0.12171649187803268, + "learning_rate": 6.409112427478281e-06, + "loss": 0.013, + "step": 85130 + }, + { + "epoch": 24.77160314227524, + "grad_norm": 0.12570516765117645, + "learning_rate": 6.4057241865001185e-06, + "loss": 0.0124, + "step": 85140 + }, + { + "epoch": 24.774512656386385, + "grad_norm": 0.08737924695014954, + "learning_rate": 6.402336598200476e-06, + "loss": 0.0113, + "step": 85150 + }, + { + "epoch": 24.777422170497527, + "grad_norm": 0.12750956416130066, + "learning_rate": 6.398949662836642e-06, + "loss": 0.0122, + "step": 85160 + }, + { + "epoch": 24.78033168460867, + "grad_norm": 0.09835779666900635, + "learning_rate": 6.395563380665818e-06, + "loss": 0.01, + "step": 85170 + }, + { + "epoch": 24.783241198719814, + "grad_norm": 0.150889590382576, + "learning_rate": 6.392177751945165e-06, + "loss": 0.0111, + "step": 85180 + }, + { + "epoch": 24.786150712830956, + "grad_norm": 0.08370453119277954, + "learning_rate": 6.388792776931806e-06, + "loss": 0.0129, + "step": 85190 + }, + { + "epoch": 24.7890602269421, + "grad_norm": 0.09563086181879044, + "learning_rate": 6.385408455882811e-06, + "loss": 0.0122, + "step": 85200 + }, + { + "epoch": 24.791969741053244, + "grad_norm": 0.12479813396930695, + "learning_rate": 6.382024789055194e-06, + "loss": 0.0119, + "step": 85210 + }, + { + "epoch": 24.79487925516439, + "grad_norm": 0.10720158368349075, + "learning_rate": 6.378641776705912e-06, + "loss": 0.012, + "step": 85220 + }, + { + "epoch": 24.79778876927553, + "grad_norm": 0.07506361603736877, + "learning_rate": 6.375259419091894e-06, + "loss": 0.0107, + "step": 85230 + }, + { + "epoch": 24.800698283386673, + "grad_norm": 0.0850244089961052, + "learning_rate": 6.371877716470009e-06, + "loss": 0.0114, + "step": 85240 + }, + { + "epoch": 24.80360779749782, + "grad_norm": 0.13011720776557922, + "learning_rate": 6.368496669097066e-06, + "loss": 0.0131, + "step": 85250 + }, + { + "epoch": 24.80651731160896, + "grad_norm": 0.14695890247821808, + "learning_rate": 6.365116277229836e-06, + "loss": 0.0127, + "step": 85260 + }, + { + "epoch": 24.809426825720106, + "grad_norm": 0.10990049690008163, + "learning_rate": 6.361736541125047e-06, + "loss": 0.0136, + "step": 85270 + }, + { + "epoch": 24.812336339831248, + "grad_norm": 0.10284925252199173, + "learning_rate": 6.35835746103936e-06, + "loss": 0.0126, + "step": 85280 + }, + { + "epoch": 24.815245853942393, + "grad_norm": 0.10569377988576889, + "learning_rate": 6.354979037229389e-06, + "loss": 0.0125, + "step": 85290 + }, + { + "epoch": 24.818155368053535, + "grad_norm": 0.12465944141149521, + "learning_rate": 6.3516012699517065e-06, + "loss": 0.0124, + "step": 85300 + }, + { + "epoch": 24.821064882164677, + "grad_norm": 0.09027573466300964, + "learning_rate": 6.348224159462839e-06, + "loss": 0.0143, + "step": 85310 + }, + { + "epoch": 24.823974396275823, + "grad_norm": 0.1031322032213211, + "learning_rate": 6.3448477060192444e-06, + "loss": 0.0134, + "step": 85320 + }, + { + "epoch": 24.826883910386965, + "grad_norm": 0.10984901338815689, + "learning_rate": 6.341471909877353e-06, + "loss": 0.0127, + "step": 85330 + }, + { + "epoch": 24.82979342449811, + "grad_norm": 0.1103210374712944, + "learning_rate": 6.338096771293524e-06, + "loss": 0.0134, + "step": 85340 + }, + { + "epoch": 24.832702938609252, + "grad_norm": 0.10986834019422531, + "learning_rate": 6.334722290524081e-06, + "loss": 0.0117, + "step": 85350 + }, + { + "epoch": 24.835612452720397, + "grad_norm": 0.07670003175735474, + "learning_rate": 6.331348467825299e-06, + "loss": 0.0122, + "step": 85360 + }, + { + "epoch": 24.83852196683154, + "grad_norm": 0.11833784729242325, + "learning_rate": 6.327975303453388e-06, + "loss": 0.0125, + "step": 85370 + }, + { + "epoch": 24.84143148094268, + "grad_norm": 0.09652166813611984, + "learning_rate": 6.324602797664527e-06, + "loss": 0.0112, + "step": 85380 + }, + { + "epoch": 24.844340995053827, + "grad_norm": 0.10168454051017761, + "learning_rate": 6.321230950714828e-06, + "loss": 0.0124, + "step": 85390 + }, + { + "epoch": 24.84725050916497, + "grad_norm": 0.09591955691576004, + "learning_rate": 6.3178597628603666e-06, + "loss": 0.0102, + "step": 85400 + }, + { + "epoch": 24.850160023276114, + "grad_norm": 0.09037522226572037, + "learning_rate": 6.314489234357154e-06, + "loss": 0.0127, + "step": 85410 + }, + { + "epoch": 24.853069537387256, + "grad_norm": 0.07726765424013138, + "learning_rate": 6.311119365461163e-06, + "loss": 0.0115, + "step": 85420 + }, + { + "epoch": 24.8559790514984, + "grad_norm": 0.1147426962852478, + "learning_rate": 6.307750156428322e-06, + "loss": 0.0114, + "step": 85430 + }, + { + "epoch": 24.858888565609544, + "grad_norm": 0.1611948311328888, + "learning_rate": 6.304381607514493e-06, + "loss": 0.0122, + "step": 85440 + }, + { + "epoch": 24.861798079720685, + "grad_norm": 0.12793035805225372, + "learning_rate": 6.301013718975483e-06, + "loss": 0.0132, + "step": 85450 + }, + { + "epoch": 24.86470759383183, + "grad_norm": 0.16777218878269196, + "learning_rate": 6.297646491067085e-06, + "loss": 0.0123, + "step": 85460 + }, + { + "epoch": 24.867617107942973, + "grad_norm": 0.11690624058246613, + "learning_rate": 6.294279924045003e-06, + "loss": 0.0122, + "step": 85470 + }, + { + "epoch": 24.87052662205412, + "grad_norm": 0.10180002450942993, + "learning_rate": 6.290914018164905e-06, + "loss": 0.012, + "step": 85480 + }, + { + "epoch": 24.87343613616526, + "grad_norm": 0.10044217854738235, + "learning_rate": 6.287548773682411e-06, + "loss": 0.0137, + "step": 85490 + }, + { + "epoch": 24.876345650276402, + "grad_norm": 0.20322756469249725, + "learning_rate": 6.2841841908530955e-06, + "loss": 0.012, + "step": 85500 + }, + { + "epoch": 24.879255164387548, + "grad_norm": 0.058610253036022186, + "learning_rate": 6.280820269932472e-06, + "loss": 0.0106, + "step": 85510 + }, + { + "epoch": 24.88216467849869, + "grad_norm": 0.10222066938877106, + "learning_rate": 6.277457011176003e-06, + "loss": 0.0117, + "step": 85520 + }, + { + "epoch": 24.885074192609835, + "grad_norm": 0.09049922227859497, + "learning_rate": 6.274094414839109e-06, + "loss": 0.0107, + "step": 85530 + }, + { + "epoch": 24.887983706720977, + "grad_norm": 0.0967082679271698, + "learning_rate": 6.270732481177165e-06, + "loss": 0.0133, + "step": 85540 + }, + { + "epoch": 24.890893220832123, + "grad_norm": 0.0962807759642601, + "learning_rate": 6.267371210445473e-06, + "loss": 0.0105, + "step": 85550 + }, + { + "epoch": 24.893802734943264, + "grad_norm": 0.11145221441984177, + "learning_rate": 6.26401060289931e-06, + "loss": 0.0123, + "step": 85560 + }, + { + "epoch": 24.896712249054406, + "grad_norm": 0.10226771980524063, + "learning_rate": 6.260650658793892e-06, + "loss": 0.0125, + "step": 85570 + }, + { + "epoch": 24.899621763165552, + "grad_norm": 0.15415678918361664, + "learning_rate": 6.2572913783843845e-06, + "loss": 0.0136, + "step": 85580 + }, + { + "epoch": 24.902531277276694, + "grad_norm": 0.11573503911495209, + "learning_rate": 6.253932761925893e-06, + "loss": 0.0116, + "step": 85590 + }, + { + "epoch": 24.90544079138784, + "grad_norm": 0.1452203392982483, + "learning_rate": 6.250574809673489e-06, + "loss": 0.0102, + "step": 85600 + }, + { + "epoch": 24.90835030549898, + "grad_norm": 0.14916515350341797, + "learning_rate": 6.247217521882193e-06, + "loss": 0.0126, + "step": 85610 + }, + { + "epoch": 24.911259819610127, + "grad_norm": 0.10076643526554108, + "learning_rate": 6.243860898806956e-06, + "loss": 0.013, + "step": 85620 + }, + { + "epoch": 24.91416933372127, + "grad_norm": 0.10956788063049316, + "learning_rate": 6.240504940702706e-06, + "loss": 0.0124, + "step": 85630 + }, + { + "epoch": 24.91707884783241, + "grad_norm": 0.10100875794887543, + "learning_rate": 6.23714964782429e-06, + "loss": 0.0117, + "step": 85640 + }, + { + "epoch": 24.919988361943556, + "grad_norm": 0.15398548543453217, + "learning_rate": 6.2337950204265316e-06, + "loss": 0.0131, + "step": 85650 + }, + { + "epoch": 24.922897876054698, + "grad_norm": 0.10184542834758759, + "learning_rate": 6.230441058764193e-06, + "loss": 0.0134, + "step": 85660 + }, + { + "epoch": 24.925807390165843, + "grad_norm": 0.13151587545871735, + "learning_rate": 6.227087763091976e-06, + "loss": 0.0102, + "step": 85670 + }, + { + "epoch": 24.928716904276985, + "grad_norm": 0.10899757593870163, + "learning_rate": 6.223735133664553e-06, + "loss": 0.012, + "step": 85680 + }, + { + "epoch": 24.93162641838813, + "grad_norm": 0.18516609072685242, + "learning_rate": 6.220383170736524e-06, + "loss": 0.0098, + "step": 85690 + }, + { + "epoch": 24.934535932499273, + "grad_norm": 0.09418153017759323, + "learning_rate": 6.217031874562458e-06, + "loss": 0.0122, + "step": 85700 + }, + { + "epoch": 24.937445446610415, + "grad_norm": 0.09670817852020264, + "learning_rate": 6.213681245396853e-06, + "loss": 0.0114, + "step": 85710 + }, + { + "epoch": 24.94035496072156, + "grad_norm": 0.13291838765144348, + "learning_rate": 6.210331283494173e-06, + "loss": 0.011, + "step": 85720 + }, + { + "epoch": 24.943264474832702, + "grad_norm": 0.11332960426807404, + "learning_rate": 6.206981989108829e-06, + "loss": 0.0112, + "step": 85730 + }, + { + "epoch": 24.946173988943848, + "grad_norm": 0.09275561571121216, + "learning_rate": 6.203633362495175e-06, + "loss": 0.0126, + "step": 85740 + }, + { + "epoch": 24.94908350305499, + "grad_norm": 0.1205025166273117, + "learning_rate": 6.2002854039075105e-06, + "loss": 0.011, + "step": 85750 + }, + { + "epoch": 24.95199301716613, + "grad_norm": 0.13624876737594604, + "learning_rate": 6.196938113600098e-06, + "loss": 0.013, + "step": 85760 + }, + { + "epoch": 24.954902531277277, + "grad_norm": 0.13278517127037048, + "learning_rate": 6.1935914918271455e-06, + "loss": 0.0118, + "step": 85770 + }, + { + "epoch": 24.95781204538842, + "grad_norm": 0.09846209734678268, + "learning_rate": 6.190245538842794e-06, + "loss": 0.0105, + "step": 85780 + }, + { + "epoch": 24.960721559499564, + "grad_norm": 0.11793617159128189, + "learning_rate": 6.186900254901158e-06, + "loss": 0.013, + "step": 85790 + }, + { + "epoch": 24.963631073610706, + "grad_norm": 0.11861302703619003, + "learning_rate": 6.183555640256289e-06, + "loss": 0.0136, + "step": 85800 + }, + { + "epoch": 24.966540587721852, + "grad_norm": 0.09886655956506729, + "learning_rate": 6.180211695162188e-06, + "loss": 0.0117, + "step": 85810 + }, + { + "epoch": 24.969450101832994, + "grad_norm": 0.12021879851818085, + "learning_rate": 6.1768684198727955e-06, + "loss": 0.0154, + "step": 85820 + }, + { + "epoch": 24.972359615944136, + "grad_norm": 0.11797337234020233, + "learning_rate": 6.173525814642021e-06, + "loss": 0.0131, + "step": 85830 + }, + { + "epoch": 24.97526913005528, + "grad_norm": 0.11607497930526733, + "learning_rate": 6.1701838797237146e-06, + "loss": 0.013, + "step": 85840 + }, + { + "epoch": 24.978178644166423, + "grad_norm": 0.13021092116832733, + "learning_rate": 6.166842615371672e-06, + "loss": 0.0109, + "step": 85850 + }, + { + "epoch": 24.98108815827757, + "grad_norm": 0.13061492145061493, + "learning_rate": 6.163502021839628e-06, + "loss": 0.0099, + "step": 85860 + }, + { + "epoch": 24.98399767238871, + "grad_norm": 0.13041238486766815, + "learning_rate": 6.160162099381301e-06, + "loss": 0.0107, + "step": 85870 + }, + { + "epoch": 24.986907186499856, + "grad_norm": 0.06641504168510437, + "learning_rate": 6.156822848250324e-06, + "loss": 0.0109, + "step": 85880 + }, + { + "epoch": 24.989816700610998, + "grad_norm": 0.08063560724258423, + "learning_rate": 6.153484268700286e-06, + "loss": 0.0114, + "step": 85890 + }, + { + "epoch": 24.99272621472214, + "grad_norm": 0.09768492728471756, + "learning_rate": 6.150146360984738e-06, + "loss": 0.0119, + "step": 85900 + }, + { + "epoch": 24.995635728833285, + "grad_norm": 0.09927854686975479, + "learning_rate": 6.146809125357173e-06, + "loss": 0.0119, + "step": 85910 + }, + { + "epoch": 24.998545242944427, + "grad_norm": 0.0972532257437706, + "learning_rate": 6.1434725620710245e-06, + "loss": 0.0123, + "step": 85920 + }, + { + "epoch": 25.001454757055573, + "grad_norm": 0.11746270954608917, + "learning_rate": 6.140136671379692e-06, + "loss": 0.0107, + "step": 85930 + }, + { + "epoch": 25.004364271166715, + "grad_norm": 0.1110803559422493, + "learning_rate": 6.136801453536503e-06, + "loss": 0.0127, + "step": 85940 + }, + { + "epoch": 25.00727378527786, + "grad_norm": 0.08417966961860657, + "learning_rate": 6.133466908794758e-06, + "loss": 0.012, + "step": 85950 + }, + { + "epoch": 25.010183299389002, + "grad_norm": 0.10145967453718185, + "learning_rate": 6.130133037407683e-06, + "loss": 0.0129, + "step": 85960 + }, + { + "epoch": 25.013092813500144, + "grad_norm": 0.14896011352539062, + "learning_rate": 6.126799839628466e-06, + "loss": 0.0113, + "step": 85970 + }, + { + "epoch": 25.01600232761129, + "grad_norm": 0.12887533009052277, + "learning_rate": 6.123467315710249e-06, + "loss": 0.0119, + "step": 85980 + }, + { + "epoch": 25.01891184172243, + "grad_norm": 0.17322488129138947, + "learning_rate": 6.1201354659061045e-06, + "loss": 0.012, + "step": 85990 + }, + { + "epoch": 25.021821355833577, + "grad_norm": 0.19435974955558777, + "learning_rate": 6.116804290469075e-06, + "loss": 0.0142, + "step": 86000 + }, + { + "epoch": 25.02473086994472, + "grad_norm": 0.11961536854505539, + "learning_rate": 6.1134737896521295e-06, + "loss": 0.0125, + "step": 86010 + }, + { + "epoch": 25.027640384055864, + "grad_norm": 0.0866071805357933, + "learning_rate": 6.110143963708203e-06, + "loss": 0.0113, + "step": 86020 + }, + { + "epoch": 25.030549898167006, + "grad_norm": 0.0879877582192421, + "learning_rate": 6.106814812890182e-06, + "loss": 0.0104, + "step": 86030 + }, + { + "epoch": 25.033459412278148, + "grad_norm": 0.07898099720478058, + "learning_rate": 6.103486337450884e-06, + "loss": 0.0115, + "step": 86040 + }, + { + "epoch": 25.036368926389294, + "grad_norm": 0.12815001606941223, + "learning_rate": 6.100158537643081e-06, + "loss": 0.0116, + "step": 86050 + }, + { + "epoch": 25.039278440500436, + "grad_norm": 0.11045181006193161, + "learning_rate": 6.0968314137195055e-06, + "loss": 0.0125, + "step": 86060 + }, + { + "epoch": 25.04218795461158, + "grad_norm": 0.08950744569301605, + "learning_rate": 6.093504965932832e-06, + "loss": 0.0113, + "step": 86070 + }, + { + "epoch": 25.045097468722723, + "grad_norm": 0.09602786600589752, + "learning_rate": 6.090179194535674e-06, + "loss": 0.0125, + "step": 86080 + }, + { + "epoch": 25.04800698283387, + "grad_norm": 0.14716146886348724, + "learning_rate": 6.086854099780607e-06, + "loss": 0.0139, + "step": 86090 + }, + { + "epoch": 25.05091649694501, + "grad_norm": 0.08243688195943832, + "learning_rate": 6.083529681920155e-06, + "loss": 0.0141, + "step": 86100 + }, + { + "epoch": 25.053826011056152, + "grad_norm": 0.12707296013832092, + "learning_rate": 6.080205941206778e-06, + "loss": 0.0138, + "step": 86110 + }, + { + "epoch": 25.056735525167298, + "grad_norm": 0.09195158630609512, + "learning_rate": 6.076882877892891e-06, + "loss": 0.0125, + "step": 86120 + }, + { + "epoch": 25.05964503927844, + "grad_norm": 0.09495742619037628, + "learning_rate": 6.0735604922308605e-06, + "loss": 0.0131, + "step": 86130 + }, + { + "epoch": 25.062554553389585, + "grad_norm": 0.1219615563750267, + "learning_rate": 6.070238784473007e-06, + "loss": 0.0108, + "step": 86140 + }, + { + "epoch": 25.065464067500727, + "grad_norm": 0.11711146682500839, + "learning_rate": 6.066917754871587e-06, + "loss": 0.0103, + "step": 86150 + }, + { + "epoch": 25.06837358161187, + "grad_norm": 0.11943256109952927, + "learning_rate": 6.063597403678799e-06, + "loss": 0.0123, + "step": 86160 + }, + { + "epoch": 25.071283095723015, + "grad_norm": 0.11109407246112823, + "learning_rate": 6.060277731146825e-06, + "loss": 0.0133, + "step": 86170 + }, + { + "epoch": 25.074192609834157, + "grad_norm": 0.11558933556079865, + "learning_rate": 6.056958737527759e-06, + "loss": 0.012, + "step": 86180 + }, + { + "epoch": 25.077102123945302, + "grad_norm": 0.11845910549163818, + "learning_rate": 6.053640423073652e-06, + "loss": 0.0098, + "step": 86190 + }, + { + "epoch": 25.080011638056444, + "grad_norm": 0.13331936299800873, + "learning_rate": 6.050322788036515e-06, + "loss": 0.0109, + "step": 86200 + }, + { + "epoch": 25.08292115216759, + "grad_norm": 0.10883145779371262, + "learning_rate": 6.0470058326683055e-06, + "loss": 0.0141, + "step": 86210 + }, + { + "epoch": 25.08583066627873, + "grad_norm": 0.12354930490255356, + "learning_rate": 6.043689557220918e-06, + "loss": 0.0095, + "step": 86220 + }, + { + "epoch": 25.088740180389873, + "grad_norm": 0.12159726768732071, + "learning_rate": 6.040373961946196e-06, + "loss": 0.0125, + "step": 86230 + }, + { + "epoch": 25.09164969450102, + "grad_norm": 0.12891516089439392, + "learning_rate": 6.037059047095944e-06, + "loss": 0.0128, + "step": 86240 + }, + { + "epoch": 25.09455920861216, + "grad_norm": 0.08604128658771515, + "learning_rate": 6.033744812921911e-06, + "loss": 0.0118, + "step": 86250 + }, + { + "epoch": 25.097468722723306, + "grad_norm": 0.09768880158662796, + "learning_rate": 6.030431259675784e-06, + "loss": 0.0119, + "step": 86260 + }, + { + "epoch": 25.100378236834448, + "grad_norm": 0.09811443835496902, + "learning_rate": 6.027118387609208e-06, + "loss": 0.0125, + "step": 86270 + }, + { + "epoch": 25.103287750945594, + "grad_norm": 0.07936377078294754, + "learning_rate": 6.02380619697378e-06, + "loss": 0.0125, + "step": 86280 + }, + { + "epoch": 25.106197265056736, + "grad_norm": 0.08743051439523697, + "learning_rate": 6.020494688021028e-06, + "loss": 0.0107, + "step": 86290 + }, + { + "epoch": 25.109106779167877, + "grad_norm": 0.14093996584415436, + "learning_rate": 6.017183861002452e-06, + "loss": 0.0114, + "step": 86300 + }, + { + "epoch": 25.112016293279023, + "grad_norm": 0.11149803549051285, + "learning_rate": 6.013873716169473e-06, + "loss": 0.0134, + "step": 86310 + }, + { + "epoch": 25.114925807390165, + "grad_norm": 0.10292930155992508, + "learning_rate": 6.010564253773486e-06, + "loss": 0.0102, + "step": 86320 + }, + { + "epoch": 25.11783532150131, + "grad_norm": 0.11240074038505554, + "learning_rate": 6.0072554740658154e-06, + "loss": 0.0131, + "step": 86330 + }, + { + "epoch": 25.120744835612452, + "grad_norm": 0.14069773256778717, + "learning_rate": 6.003947377297751e-06, + "loss": 0.0121, + "step": 86340 + }, + { + "epoch": 25.123654349723598, + "grad_norm": 0.13305328786373138, + "learning_rate": 6.000639963720507e-06, + "loss": 0.0126, + "step": 86350 + }, + { + "epoch": 25.12656386383474, + "grad_norm": 0.1065889298915863, + "learning_rate": 5.997333233585268e-06, + "loss": 0.0134, + "step": 86360 + }, + { + "epoch": 25.12947337794588, + "grad_norm": 0.0846293717622757, + "learning_rate": 5.994027187143163e-06, + "loss": 0.0104, + "step": 86370 + }, + { + "epoch": 25.132382892057027, + "grad_norm": 0.15310800075531006, + "learning_rate": 5.990721824645254e-06, + "loss": 0.0126, + "step": 86380 + }, + { + "epoch": 25.13529240616817, + "grad_norm": 0.10960552841424942, + "learning_rate": 5.987417146342565e-06, + "loss": 0.0114, + "step": 86390 + }, + { + "epoch": 25.138201920279315, + "grad_norm": 0.0988236665725708, + "learning_rate": 5.9841131524860735e-06, + "loss": 0.0118, + "step": 86400 + }, + { + "epoch": 25.141111434390456, + "grad_norm": 0.1094384491443634, + "learning_rate": 5.980809843326685e-06, + "loss": 0.0124, + "step": 86410 + }, + { + "epoch": 25.144020948501602, + "grad_norm": 0.11255045235157013, + "learning_rate": 5.977507219115265e-06, + "loss": 0.012, + "step": 86420 + }, + { + "epoch": 25.146930462612744, + "grad_norm": 0.11158303171396255, + "learning_rate": 5.9742052801026265e-06, + "loss": 0.0126, + "step": 86430 + }, + { + "epoch": 25.149839976723886, + "grad_norm": 0.08343951404094696, + "learning_rate": 5.970904026539538e-06, + "loss": 0.0136, + "step": 86440 + }, + { + "epoch": 25.15274949083503, + "grad_norm": 0.12121092528104782, + "learning_rate": 5.967603458676703e-06, + "loss": 0.0116, + "step": 86450 + }, + { + "epoch": 25.155659004946173, + "grad_norm": 0.08739136904478073, + "learning_rate": 5.964303576764764e-06, + "loss": 0.0104, + "step": 86460 + }, + { + "epoch": 25.15856851905732, + "grad_norm": 0.10387146472930908, + "learning_rate": 5.961004381054351e-06, + "loss": 0.0146, + "step": 86470 + }, + { + "epoch": 25.16147803316846, + "grad_norm": 0.10584622621536255, + "learning_rate": 5.957705871796001e-06, + "loss": 0.0148, + "step": 86480 + }, + { + "epoch": 25.164387547279603, + "grad_norm": 0.11739082634449005, + "learning_rate": 5.954408049240211e-06, + "loss": 0.0126, + "step": 86490 + }, + { + "epoch": 25.167297061390748, + "grad_norm": 0.09067046642303467, + "learning_rate": 5.951110913637435e-06, + "loss": 0.0107, + "step": 86500 + }, + { + "epoch": 25.17020657550189, + "grad_norm": 0.101159006357193, + "learning_rate": 5.9478144652380725e-06, + "loss": 0.0128, + "step": 86510 + }, + { + "epoch": 25.173116089613035, + "grad_norm": 0.09772844612598419, + "learning_rate": 5.9445187042924615e-06, + "loss": 0.0123, + "step": 86520 + }, + { + "epoch": 25.176025603724177, + "grad_norm": 0.14358943700790405, + "learning_rate": 5.9412236310508896e-06, + "loss": 0.0137, + "step": 86530 + }, + { + "epoch": 25.178935117835323, + "grad_norm": 0.10577183216810226, + "learning_rate": 5.937929245763599e-06, + "loss": 0.0128, + "step": 86540 + }, + { + "epoch": 25.181844631946465, + "grad_norm": 0.13220950961112976, + "learning_rate": 5.934635548680784e-06, + "loss": 0.0134, + "step": 86550 + }, + { + "epoch": 25.184754146057607, + "grad_norm": 0.07032065838575363, + "learning_rate": 5.931342540052567e-06, + "loss": 0.012, + "step": 86560 + }, + { + "epoch": 25.187663660168752, + "grad_norm": 0.13498838245868683, + "learning_rate": 5.928050220129037e-06, + "loss": 0.0123, + "step": 86570 + }, + { + "epoch": 25.190573174279894, + "grad_norm": 0.12329194694757462, + "learning_rate": 5.9247585891602265e-06, + "loss": 0.0125, + "step": 86580 + }, + { + "epoch": 25.19348268839104, + "grad_norm": 0.11143108457326889, + "learning_rate": 5.921467647396112e-06, + "loss": 0.0127, + "step": 86590 + }, + { + "epoch": 25.19639220250218, + "grad_norm": 0.10464686155319214, + "learning_rate": 5.9181773950866114e-06, + "loss": 0.0117, + "step": 86600 + }, + { + "epoch": 25.199301716613327, + "grad_norm": 0.07959343492984772, + "learning_rate": 5.9148878324816016e-06, + "loss": 0.0129, + "step": 86610 + }, + { + "epoch": 25.20221123072447, + "grad_norm": 0.1128874197602272, + "learning_rate": 5.91159895983091e-06, + "loss": 0.0123, + "step": 86620 + }, + { + "epoch": 25.20512074483561, + "grad_norm": 0.11910408735275269, + "learning_rate": 5.908310777384292e-06, + "loss": 0.0114, + "step": 86630 + }, + { + "epoch": 25.208030258946756, + "grad_norm": 0.10409354418516159, + "learning_rate": 5.905023285391479e-06, + "loss": 0.0112, + "step": 86640 + }, + { + "epoch": 25.2109397730579, + "grad_norm": 0.12335251271724701, + "learning_rate": 5.901736484102119e-06, + "loss": 0.012, + "step": 86650 + }, + { + "epoch": 25.213849287169044, + "grad_norm": 0.11148657649755478, + "learning_rate": 5.898450373765829e-06, + "loss": 0.012, + "step": 86660 + }, + { + "epoch": 25.216758801280186, + "grad_norm": 0.10629821568727493, + "learning_rate": 5.895164954632176e-06, + "loss": 0.0096, + "step": 86670 + }, + { + "epoch": 25.21966831539133, + "grad_norm": 0.12796781957149506, + "learning_rate": 5.891880226950651e-06, + "loss": 0.0116, + "step": 86680 + }, + { + "epoch": 25.222577829502473, + "grad_norm": 0.09295260906219482, + "learning_rate": 5.88859619097072e-06, + "loss": 0.0106, + "step": 86690 + }, + { + "epoch": 25.225487343613615, + "grad_norm": 0.1172393187880516, + "learning_rate": 5.885312846941772e-06, + "loss": 0.0122, + "step": 86700 + }, + { + "epoch": 25.22839685772476, + "grad_norm": 0.11674809455871582, + "learning_rate": 5.882030195113166e-06, + "loss": 0.0116, + "step": 86710 + }, + { + "epoch": 25.231306371835903, + "grad_norm": 0.0941578820347786, + "learning_rate": 5.87874823573419e-06, + "loss": 0.0109, + "step": 86720 + }, + { + "epoch": 25.234215885947048, + "grad_norm": 0.10578205436468124, + "learning_rate": 5.8754669690540905e-06, + "loss": 0.0127, + "step": 86730 + }, + { + "epoch": 25.23712540005819, + "grad_norm": 0.0962415337562561, + "learning_rate": 5.872186395322063e-06, + "loss": 0.011, + "step": 86740 + }, + { + "epoch": 25.240034914169335, + "grad_norm": 0.11222007870674133, + "learning_rate": 5.8689065147872404e-06, + "loss": 0.0113, + "step": 86750 + }, + { + "epoch": 25.242944428280477, + "grad_norm": 0.10162713378667831, + "learning_rate": 5.865627327698698e-06, + "loss": 0.0112, + "step": 86760 + }, + { + "epoch": 25.24585394239162, + "grad_norm": 0.1168115884065628, + "learning_rate": 5.862348834305489e-06, + "loss": 0.0131, + "step": 86770 + }, + { + "epoch": 25.248763456502765, + "grad_norm": 0.07431214302778244, + "learning_rate": 5.859071034856582e-06, + "loss": 0.0133, + "step": 86780 + }, + { + "epoch": 25.251672970613907, + "grad_norm": 0.1111876592040062, + "learning_rate": 5.8557939296009e-06, + "loss": 0.0121, + "step": 86790 + }, + { + "epoch": 25.254582484725052, + "grad_norm": 0.09611357748508453, + "learning_rate": 5.852517518787323e-06, + "loss": 0.0113, + "step": 86800 + }, + { + "epoch": 25.257491998836194, + "grad_norm": 0.09273262321949005, + "learning_rate": 5.8492418026646766e-06, + "loss": 0.0117, + "step": 86810 + }, + { + "epoch": 25.26040151294734, + "grad_norm": 0.1644279956817627, + "learning_rate": 5.845966781481725e-06, + "loss": 0.0132, + "step": 86820 + }, + { + "epoch": 25.26331102705848, + "grad_norm": 0.1441599726676941, + "learning_rate": 5.842692455487181e-06, + "loss": 0.0136, + "step": 86830 + }, + { + "epoch": 25.266220541169623, + "grad_norm": 0.2767721116542816, + "learning_rate": 5.839418824929713e-06, + "loss": 0.012, + "step": 86840 + }, + { + "epoch": 25.26913005528077, + "grad_norm": 0.1296311616897583, + "learning_rate": 5.836145890057933e-06, + "loss": 0.0134, + "step": 86850 + }, + { + "epoch": 25.27203956939191, + "grad_norm": 0.0798201784491539, + "learning_rate": 5.832873651120394e-06, + "loss": 0.0106, + "step": 86860 + }, + { + "epoch": 25.274949083503056, + "grad_norm": 0.1196063980460167, + "learning_rate": 5.829602108365602e-06, + "loss": 0.0127, + "step": 86870 + }, + { + "epoch": 25.2778585976142, + "grad_norm": 0.08389527350664139, + "learning_rate": 5.826331262042015e-06, + "loss": 0.0107, + "step": 86880 + }, + { + "epoch": 25.28076811172534, + "grad_norm": 0.08662257343530655, + "learning_rate": 5.823061112398031e-06, + "loss": 0.0115, + "step": 86890 + }, + { + "epoch": 25.283677625836486, + "grad_norm": 0.08979322016239166, + "learning_rate": 5.8197916596819844e-06, + "loss": 0.0109, + "step": 86900 + }, + { + "epoch": 25.286587139947628, + "grad_norm": 0.1421636939048767, + "learning_rate": 5.816522904142179e-06, + "loss": 0.0138, + "step": 86910 + }, + { + "epoch": 25.289496654058773, + "grad_norm": 0.13639117777347565, + "learning_rate": 5.8132548460268575e-06, + "loss": 0.0125, + "step": 86920 + }, + { + "epoch": 25.292406168169915, + "grad_norm": 0.1142483577132225, + "learning_rate": 5.8099874855841974e-06, + "loss": 0.0115, + "step": 86930 + }, + { + "epoch": 25.29531568228106, + "grad_norm": 0.09981069713830948, + "learning_rate": 5.8067208230623455e-06, + "loss": 0.0133, + "step": 86940 + }, + { + "epoch": 25.298225196392202, + "grad_norm": 0.17936305701732635, + "learning_rate": 5.803454858709371e-06, + "loss": 0.0111, + "step": 86950 + }, + { + "epoch": 25.301134710503344, + "grad_norm": 0.12150456011295319, + "learning_rate": 5.8001895927733135e-06, + "loss": 0.0121, + "step": 86960 + }, + { + "epoch": 25.30404422461449, + "grad_norm": 0.14500518143177032, + "learning_rate": 5.796925025502138e-06, + "loss": 0.0149, + "step": 86970 + }, + { + "epoch": 25.306953738725632, + "grad_norm": 0.12600308656692505, + "learning_rate": 5.793661157143772e-06, + "loss": 0.0113, + "step": 86980 + }, + { + "epoch": 25.309863252836777, + "grad_norm": 0.12201393395662308, + "learning_rate": 5.790397987946088e-06, + "loss": 0.0122, + "step": 86990 + }, + { + "epoch": 25.31277276694792, + "grad_norm": 0.08881831914186478, + "learning_rate": 5.787135518156895e-06, + "loss": 0.0143, + "step": 87000 + }, + { + "epoch": 25.315682281059065, + "grad_norm": 0.10068480670452118, + "learning_rate": 5.783873748023966e-06, + "loss": 0.013, + "step": 87010 + }, + { + "epoch": 25.318591795170207, + "grad_norm": 0.08728103339672089, + "learning_rate": 5.7806126777949975e-06, + "loss": 0.0112, + "step": 87020 + }, + { + "epoch": 25.32150130928135, + "grad_norm": 0.10652013123035431, + "learning_rate": 5.7773523077176544e-06, + "loss": 0.0114, + "step": 87030 + }, + { + "epoch": 25.324410823392494, + "grad_norm": 0.12219797819852829, + "learning_rate": 5.774092638039543e-06, + "loss": 0.0132, + "step": 87040 + }, + { + "epoch": 25.327320337503636, + "grad_norm": 0.08915212750434875, + "learning_rate": 5.770833669008213e-06, + "loss": 0.0102, + "step": 87050 + }, + { + "epoch": 25.33022985161478, + "grad_norm": 0.12581028044223785, + "learning_rate": 5.7675754008711515e-06, + "loss": 0.0138, + "step": 87060 + }, + { + "epoch": 25.333139365725923, + "grad_norm": 0.12684214115142822, + "learning_rate": 5.76431783387581e-06, + "loss": 0.0136, + "step": 87070 + }, + { + "epoch": 25.33604887983707, + "grad_norm": 0.08699925988912582, + "learning_rate": 5.761060968269583e-06, + "loss": 0.0126, + "step": 87080 + }, + { + "epoch": 25.33895839394821, + "grad_norm": 0.11143616586923599, + "learning_rate": 5.757804804299801e-06, + "loss": 0.0092, + "step": 87090 + }, + { + "epoch": 25.341867908059353, + "grad_norm": 0.13559377193450928, + "learning_rate": 5.754549342213749e-06, + "loss": 0.0154, + "step": 87100 + }, + { + "epoch": 25.3447774221705, + "grad_norm": 0.0945858359336853, + "learning_rate": 5.751294582258665e-06, + "loss": 0.0124, + "step": 87110 + }, + { + "epoch": 25.34768693628164, + "grad_norm": 0.09698645025491714, + "learning_rate": 5.748040524681722e-06, + "loss": 0.0134, + "step": 87120 + }, + { + "epoch": 25.350596450392786, + "grad_norm": 0.1870017945766449, + "learning_rate": 5.744787169730037e-06, + "loss": 0.0111, + "step": 87130 + }, + { + "epoch": 25.353505964503928, + "grad_norm": 0.13926786184310913, + "learning_rate": 5.7415345176506905e-06, + "loss": 0.0107, + "step": 87140 + }, + { + "epoch": 25.35641547861507, + "grad_norm": 0.07746720314025879, + "learning_rate": 5.738282568690699e-06, + "loss": 0.0119, + "step": 87150 + }, + { + "epoch": 25.359324992726215, + "grad_norm": 0.11385714262723923, + "learning_rate": 5.7350313230970255e-06, + "loss": 0.0105, + "step": 87160 + }, + { + "epoch": 25.362234506837357, + "grad_norm": 0.07208363711833954, + "learning_rate": 5.731780781116569e-06, + "loss": 0.0111, + "step": 87170 + }, + { + "epoch": 25.365144020948502, + "grad_norm": 0.11877432465553284, + "learning_rate": 5.728530942996207e-06, + "loss": 0.0131, + "step": 87180 + }, + { + "epoch": 25.368053535059644, + "grad_norm": 0.10301011055707932, + "learning_rate": 5.725281808982735e-06, + "loss": 0.0137, + "step": 87190 + }, + { + "epoch": 25.37096304917079, + "grad_norm": 0.08781221508979797, + "learning_rate": 5.722033379322895e-06, + "loss": 0.0118, + "step": 87200 + }, + { + "epoch": 25.37387256328193, + "grad_norm": 0.26018327474594116, + "learning_rate": 5.718785654263392e-06, + "loss": 0.0123, + "step": 87210 + }, + { + "epoch": 25.376782077393074, + "grad_norm": 0.11429275572299957, + "learning_rate": 5.715538634050872e-06, + "loss": 0.0117, + "step": 87220 + }, + { + "epoch": 25.37969159150422, + "grad_norm": 0.16038130223751068, + "learning_rate": 5.712292318931916e-06, + "loss": 0.0125, + "step": 87230 + }, + { + "epoch": 25.38260110561536, + "grad_norm": 0.11496246606111526, + "learning_rate": 5.70904670915307e-06, + "loss": 0.0124, + "step": 87240 + }, + { + "epoch": 25.385510619726507, + "grad_norm": 0.09996509552001953, + "learning_rate": 5.705801804960805e-06, + "loss": 0.0128, + "step": 87250 + }, + { + "epoch": 25.38842013383765, + "grad_norm": 0.11457006633281708, + "learning_rate": 5.702557606601564e-06, + "loss": 0.0137, + "step": 87260 + }, + { + "epoch": 25.391329647948794, + "grad_norm": 0.12397462874650955, + "learning_rate": 5.699314114321709e-06, + "loss": 0.0136, + "step": 87270 + }, + { + "epoch": 25.394239162059936, + "grad_norm": 0.11455778777599335, + "learning_rate": 5.696071328367568e-06, + "loss": 0.0147, + "step": 87280 + }, + { + "epoch": 25.397148676171078, + "grad_norm": 0.11188450455665588, + "learning_rate": 5.692829248985414e-06, + "loss": 0.0121, + "step": 87290 + }, + { + "epoch": 25.400058190282223, + "grad_norm": 0.1751265972852707, + "learning_rate": 5.689587876421451e-06, + "loss": 0.0113, + "step": 87300 + }, + { + "epoch": 25.402967704393365, + "grad_norm": 0.09795019775629044, + "learning_rate": 5.6863472109218506e-06, + "loss": 0.0138, + "step": 87310 + }, + { + "epoch": 25.40587721850451, + "grad_norm": 0.12966565787792206, + "learning_rate": 5.683107252732709e-06, + "loss": 0.0124, + "step": 87320 + }, + { + "epoch": 25.408786732615653, + "grad_norm": 0.12447944283485413, + "learning_rate": 5.679868002100092e-06, + "loss": 0.0136, + "step": 87330 + }, + { + "epoch": 25.411696246726798, + "grad_norm": 0.0785774365067482, + "learning_rate": 5.676629459269987e-06, + "loss": 0.0104, + "step": 87340 + }, + { + "epoch": 25.41460576083794, + "grad_norm": 0.09878047555685043, + "learning_rate": 5.67339162448835e-06, + "loss": 0.0132, + "step": 87350 + }, + { + "epoch": 25.417515274949082, + "grad_norm": 0.1129290759563446, + "learning_rate": 5.670154498001064e-06, + "loss": 0.0112, + "step": 87360 + }, + { + "epoch": 25.420424789060228, + "grad_norm": 0.1084652841091156, + "learning_rate": 5.666918080053974e-06, + "loss": 0.0138, + "step": 87370 + }, + { + "epoch": 25.42333430317137, + "grad_norm": 0.10602705925703049, + "learning_rate": 5.663682370892867e-06, + "loss": 0.0112, + "step": 87380 + }, + { + "epoch": 25.426243817282515, + "grad_norm": 0.1279289871454239, + "learning_rate": 5.660447370763464e-06, + "loss": 0.0134, + "step": 87390 + }, + { + "epoch": 25.429153331393657, + "grad_norm": 0.11082940548658371, + "learning_rate": 5.657213079911447e-06, + "loss": 0.0122, + "step": 87400 + }, + { + "epoch": 25.432062845504802, + "grad_norm": 0.12318835407495499, + "learning_rate": 5.653979498582445e-06, + "loss": 0.0139, + "step": 87410 + }, + { + "epoch": 25.434972359615944, + "grad_norm": 0.10859766602516174, + "learning_rate": 5.650746627022022e-06, + "loss": 0.0114, + "step": 87420 + }, + { + "epoch": 25.437881873727086, + "grad_norm": 0.08902312070131302, + "learning_rate": 5.647514465475689e-06, + "loss": 0.0128, + "step": 87430 + }, + { + "epoch": 25.44079138783823, + "grad_norm": 0.14400836825370789, + "learning_rate": 5.6442830141889094e-06, + "loss": 0.0122, + "step": 87440 + }, + { + "epoch": 25.443700901949374, + "grad_norm": 0.14861242473125458, + "learning_rate": 5.641052273407098e-06, + "loss": 0.0124, + "step": 87450 + }, + { + "epoch": 25.44661041606052, + "grad_norm": 0.122688427567482, + "learning_rate": 5.637822243375604e-06, + "loss": 0.0159, + "step": 87460 + }, + { + "epoch": 25.44951993017166, + "grad_norm": 0.10872464627027512, + "learning_rate": 5.634592924339714e-06, + "loss": 0.0116, + "step": 87470 + }, + { + "epoch": 25.452429444282807, + "grad_norm": 0.07674426585435867, + "learning_rate": 5.631364316544696e-06, + "loss": 0.012, + "step": 87480 + }, + { + "epoch": 25.45533895839395, + "grad_norm": 0.1386343538761139, + "learning_rate": 5.628136420235731e-06, + "loss": 0.01, + "step": 87490 + }, + { + "epoch": 25.45824847250509, + "grad_norm": 0.07711838185787201, + "learning_rate": 5.62490923565795e-06, + "loss": 0.0128, + "step": 87500 + }, + { + "epoch": 25.461157986616236, + "grad_norm": 0.09883510321378708, + "learning_rate": 5.621682763056443e-06, + "loss": 0.0111, + "step": 87510 + }, + { + "epoch": 25.464067500727378, + "grad_norm": 0.13884173333644867, + "learning_rate": 5.6184570026762425e-06, + "loss": 0.0144, + "step": 87520 + }, + { + "epoch": 25.466977014838523, + "grad_norm": 0.11150261759757996, + "learning_rate": 5.615231954762322e-06, + "loss": 0.0124, + "step": 87530 + }, + { + "epoch": 25.469886528949665, + "grad_norm": 0.12661266326904297, + "learning_rate": 5.612007619559594e-06, + "loss": 0.0101, + "step": 87540 + }, + { + "epoch": 25.472796043060807, + "grad_norm": 0.09870382398366928, + "learning_rate": 5.608783997312932e-06, + "loss": 0.0123, + "step": 87550 + }, + { + "epoch": 25.475705557171953, + "grad_norm": 0.09002383798360825, + "learning_rate": 5.605561088267154e-06, + "loss": 0.0116, + "step": 87560 + }, + { + "epoch": 25.478615071283095, + "grad_norm": 0.10871091485023499, + "learning_rate": 5.602338892667007e-06, + "loss": 0.0113, + "step": 87570 + }, + { + "epoch": 25.48152458539424, + "grad_norm": 0.11083795130252838, + "learning_rate": 5.599117410757203e-06, + "loss": 0.0113, + "step": 87580 + }, + { + "epoch": 25.484434099505382, + "grad_norm": 0.07435361295938492, + "learning_rate": 5.595896642782397e-06, + "loss": 0.0123, + "step": 87590 + }, + { + "epoch": 25.487343613616527, + "grad_norm": 0.15081092715263367, + "learning_rate": 5.592676588987174e-06, + "loss": 0.0137, + "step": 87600 + }, + { + "epoch": 25.49025312772767, + "grad_norm": 0.17939072847366333, + "learning_rate": 5.589457249616086e-06, + "loss": 0.0113, + "step": 87610 + }, + { + "epoch": 25.49316264183881, + "grad_norm": 0.1547721028327942, + "learning_rate": 5.58623862491361e-06, + "loss": 0.0138, + "step": 87620 + }, + { + "epoch": 25.496072155949957, + "grad_norm": 0.1277560591697693, + "learning_rate": 5.583020715124191e-06, + "loss": 0.0126, + "step": 87630 + }, + { + "epoch": 25.4989816700611, + "grad_norm": 0.13131898641586304, + "learning_rate": 5.579803520492197e-06, + "loss": 0.0127, + "step": 87640 + }, + { + "epoch": 25.501891184172244, + "grad_norm": 0.10803813487291336, + "learning_rate": 5.576587041261962e-06, + "loss": 0.0122, + "step": 87650 + }, + { + "epoch": 25.504800698283386, + "grad_norm": 0.09478063881397247, + "learning_rate": 5.573371277677748e-06, + "loss": 0.0135, + "step": 87660 + }, + { + "epoch": 25.50771021239453, + "grad_norm": 0.08231937140226364, + "learning_rate": 5.5701562299837745e-06, + "loss": 0.012, + "step": 87670 + }, + { + "epoch": 25.510619726505674, + "grad_norm": 0.11144640296697617, + "learning_rate": 5.566941898424208e-06, + "loss": 0.011, + "step": 87680 + }, + { + "epoch": 25.513529240616815, + "grad_norm": 0.11873196810483932, + "learning_rate": 5.5637282832431476e-06, + "loss": 0.0112, + "step": 87690 + }, + { + "epoch": 25.51643875472796, + "grad_norm": 0.07882092893123627, + "learning_rate": 5.56051538468465e-06, + "loss": 0.0115, + "step": 87700 + }, + { + "epoch": 25.519348268839103, + "grad_norm": 0.10871482640504837, + "learning_rate": 5.557303202992719e-06, + "loss": 0.0112, + "step": 87710 + }, + { + "epoch": 25.52225778295025, + "grad_norm": 0.10573923587799072, + "learning_rate": 5.554091738411292e-06, + "loss": 0.011, + "step": 87720 + }, + { + "epoch": 25.52516729706139, + "grad_norm": 0.07802636921405792, + "learning_rate": 5.550880991184256e-06, + "loss": 0.0117, + "step": 87730 + }, + { + "epoch": 25.528076811172536, + "grad_norm": 0.11789896339178085, + "learning_rate": 5.5476709615554514e-06, + "loss": 0.0122, + "step": 87740 + }, + { + "epoch": 25.530986325283678, + "grad_norm": 0.11713097244501114, + "learning_rate": 5.5444616497686605e-06, + "loss": 0.0098, + "step": 87750 + }, + { + "epoch": 25.53389583939482, + "grad_norm": 0.09881651401519775, + "learning_rate": 5.5412530560676065e-06, + "loss": 0.0124, + "step": 87760 + }, + { + "epoch": 25.536805353505965, + "grad_norm": 0.10137613862752914, + "learning_rate": 5.5380451806959495e-06, + "loss": 0.0127, + "step": 87770 + }, + { + "epoch": 25.539714867617107, + "grad_norm": 0.11559659987688065, + "learning_rate": 5.534838023897329e-06, + "loss": 0.0116, + "step": 87780 + }, + { + "epoch": 25.542624381728253, + "grad_norm": 0.08491960912942886, + "learning_rate": 5.531631585915296e-06, + "loss": 0.0101, + "step": 87790 + }, + { + "epoch": 25.545533895839394, + "grad_norm": 0.11502115428447723, + "learning_rate": 5.528425866993353e-06, + "loss": 0.0122, + "step": 87800 + }, + { + "epoch": 25.54844340995054, + "grad_norm": 0.06962329149246216, + "learning_rate": 5.525220867374957e-06, + "loss": 0.0117, + "step": 87810 + }, + { + "epoch": 25.551352924061682, + "grad_norm": 0.1086135283112526, + "learning_rate": 5.5220165873035156e-06, + "loss": 0.0119, + "step": 87820 + }, + { + "epoch": 25.554262438172824, + "grad_norm": 0.10104059427976608, + "learning_rate": 5.518813027022363e-06, + "loss": 0.0131, + "step": 87830 + }, + { + "epoch": 25.55717195228397, + "grad_norm": 0.1369713544845581, + "learning_rate": 5.515610186774787e-06, + "loss": 0.0101, + "step": 87840 + }, + { + "epoch": 25.56008146639511, + "grad_norm": 0.11611921340227127, + "learning_rate": 5.512408066804025e-06, + "loss": 0.0125, + "step": 87850 + }, + { + "epoch": 25.562990980506257, + "grad_norm": 0.08678068965673447, + "learning_rate": 5.509206667353264e-06, + "loss": 0.0108, + "step": 87860 + }, + { + "epoch": 25.5659004946174, + "grad_norm": 0.09598438441753387, + "learning_rate": 5.506005988665617e-06, + "loss": 0.0126, + "step": 87870 + }, + { + "epoch": 25.56881000872854, + "grad_norm": 0.09993021935224533, + "learning_rate": 5.502806030984159e-06, + "loss": 0.011, + "step": 87880 + }, + { + "epoch": 25.571719522839686, + "grad_norm": 0.12174860388040543, + "learning_rate": 5.499606794551913e-06, + "loss": 0.0112, + "step": 87890 + }, + { + "epoch": 25.574629036950828, + "grad_norm": 0.11917684227228165, + "learning_rate": 5.496408279611833e-06, + "loss": 0.0118, + "step": 87900 + }, + { + "epoch": 25.577538551061973, + "grad_norm": 0.11314687132835388, + "learning_rate": 5.493210486406823e-06, + "loss": 0.0101, + "step": 87910 + }, + { + "epoch": 25.580448065173115, + "grad_norm": 0.12913261353969574, + "learning_rate": 5.490013415179735e-06, + "loss": 0.0116, + "step": 87920 + }, + { + "epoch": 25.58335757928426, + "grad_norm": 0.11864857375621796, + "learning_rate": 5.486817066173374e-06, + "loss": 0.0104, + "step": 87930 + }, + { + "epoch": 25.586267093395403, + "grad_norm": 0.10822846740484238, + "learning_rate": 5.483621439630469e-06, + "loss": 0.0124, + "step": 87940 + }, + { + "epoch": 25.589176607506545, + "grad_norm": 0.11641058325767517, + "learning_rate": 5.480426535793719e-06, + "loss": 0.0121, + "step": 87950 + }, + { + "epoch": 25.59208612161769, + "grad_norm": 0.10689792037010193, + "learning_rate": 5.477232354905746e-06, + "loss": 0.011, + "step": 87960 + }, + { + "epoch": 25.594995635728832, + "grad_norm": 0.11325965821743011, + "learning_rate": 5.474038897209129e-06, + "loss": 0.0109, + "step": 87970 + }, + { + "epoch": 25.597905149839978, + "grad_norm": 0.09292846918106079, + "learning_rate": 5.470846162946398e-06, + "loss": 0.0118, + "step": 87980 + }, + { + "epoch": 25.60081466395112, + "grad_norm": 0.08531259000301361, + "learning_rate": 5.4676541523600075e-06, + "loss": 0.0101, + "step": 87990 + }, + { + "epoch": 25.603724178062265, + "grad_norm": 0.08686395734548569, + "learning_rate": 5.464462865692382e-06, + "loss": 0.0113, + "step": 88000 + }, + { + "epoch": 25.606633692173407, + "grad_norm": 0.11074709892272949, + "learning_rate": 5.461272303185868e-06, + "loss": 0.0131, + "step": 88010 + }, + { + "epoch": 25.60954320628455, + "grad_norm": 0.11511655896902084, + "learning_rate": 5.458082465082778e-06, + "loss": 0.0123, + "step": 88020 + }, + { + "epoch": 25.612452720395694, + "grad_norm": 0.10177662968635559, + "learning_rate": 5.454893351625348e-06, + "loss": 0.0115, + "step": 88030 + }, + { + "epoch": 25.615362234506836, + "grad_norm": 0.06893067806959152, + "learning_rate": 5.4517049630557756e-06, + "loss": 0.011, + "step": 88040 + }, + { + "epoch": 25.618271748617982, + "grad_norm": 0.0794714167714119, + "learning_rate": 5.448517299616203e-06, + "loss": 0.0107, + "step": 88050 + }, + { + "epoch": 25.621181262729124, + "grad_norm": 0.08444253355264664, + "learning_rate": 5.445330361548707e-06, + "loss": 0.0128, + "step": 88060 + }, + { + "epoch": 25.62409077684027, + "grad_norm": 0.1005539670586586, + "learning_rate": 5.442144149095304e-06, + "loss": 0.0122, + "step": 88070 + }, + { + "epoch": 25.62700029095141, + "grad_norm": 0.10637499392032623, + "learning_rate": 5.438958662497988e-06, + "loss": 0.015, + "step": 88080 + }, + { + "epoch": 25.629909805062553, + "grad_norm": 0.11277727037668228, + "learning_rate": 5.435773901998663e-06, + "loss": 0.0117, + "step": 88090 + }, + { + "epoch": 25.6328193191737, + "grad_norm": 0.08572866022586823, + "learning_rate": 5.432589867839189e-06, + "loss": 0.0131, + "step": 88100 + }, + { + "epoch": 25.63572883328484, + "grad_norm": 0.15232455730438232, + "learning_rate": 5.429406560261372e-06, + "loss": 0.0142, + "step": 88110 + }, + { + "epoch": 25.638638347395986, + "grad_norm": 0.0767337903380394, + "learning_rate": 5.4262239795069734e-06, + "loss": 0.0114, + "step": 88120 + }, + { + "epoch": 25.641547861507128, + "grad_norm": 0.12770430743694305, + "learning_rate": 5.423042125817681e-06, + "loss": 0.012, + "step": 88130 + }, + { + "epoch": 25.644457375618273, + "grad_norm": 0.09571942687034607, + "learning_rate": 5.419860999435133e-06, + "loss": 0.0109, + "step": 88140 + }, + { + "epoch": 25.647366889729415, + "grad_norm": 0.09836230427026749, + "learning_rate": 5.4166806006009175e-06, + "loss": 0.0115, + "step": 88150 + }, + { + "epoch": 25.650276403840557, + "grad_norm": 0.103660948574543, + "learning_rate": 5.413500929556572e-06, + "loss": 0.0104, + "step": 88160 + }, + { + "epoch": 25.653185917951703, + "grad_norm": 0.08697967976331711, + "learning_rate": 5.410321986543565e-06, + "loss": 0.0128, + "step": 88170 + }, + { + "epoch": 25.656095432062845, + "grad_norm": 0.14189639687538147, + "learning_rate": 5.407143771803305e-06, + "loss": 0.0095, + "step": 88180 + }, + { + "epoch": 25.65900494617399, + "grad_norm": 0.18440014123916626, + "learning_rate": 5.403966285577181e-06, + "loss": 0.0109, + "step": 88190 + }, + { + "epoch": 25.661914460285132, + "grad_norm": 0.1162438839673996, + "learning_rate": 5.4007895281064874e-06, + "loss": 0.0117, + "step": 88200 + }, + { + "epoch": 25.664823974396278, + "grad_norm": 0.09603282809257507, + "learning_rate": 5.397613499632476e-06, + "loss": 0.0124, + "step": 88210 + }, + { + "epoch": 25.66773348850742, + "grad_norm": 0.11100269109010696, + "learning_rate": 5.39443820039635e-06, + "loss": 0.0115, + "step": 88220 + }, + { + "epoch": 25.67064300261856, + "grad_norm": 0.11508617550134659, + "learning_rate": 5.391263630639255e-06, + "loss": 0.0145, + "step": 88230 + }, + { + "epoch": 25.673552516729707, + "grad_norm": 0.08719713240861893, + "learning_rate": 5.38808979060227e-06, + "loss": 0.0108, + "step": 88240 + }, + { + "epoch": 25.67646203084085, + "grad_norm": 0.12943428754806519, + "learning_rate": 5.384916680526438e-06, + "loss": 0.0127, + "step": 88250 + }, + { + "epoch": 25.679371544951994, + "grad_norm": 0.08845295757055283, + "learning_rate": 5.381744300652724e-06, + "loss": 0.0101, + "step": 88260 + }, + { + "epoch": 25.682281059063136, + "grad_norm": 0.11877479404211044, + "learning_rate": 5.37857265122206e-06, + "loss": 0.0116, + "step": 88270 + }, + { + "epoch": 25.68519057317428, + "grad_norm": 0.08659867942333221, + "learning_rate": 5.375401732475303e-06, + "loss": 0.0113, + "step": 88280 + }, + { + "epoch": 25.688100087285424, + "grad_norm": 0.08103295415639877, + "learning_rate": 5.372231544653269e-06, + "loss": 0.0132, + "step": 88290 + }, + { + "epoch": 25.691009601396566, + "grad_norm": 0.11643160134553909, + "learning_rate": 5.369062087996716e-06, + "loss": 0.014, + "step": 88300 + }, + { + "epoch": 25.69391911550771, + "grad_norm": 0.11680260300636292, + "learning_rate": 5.3658933627463334e-06, + "loss": 0.0111, + "step": 88310 + }, + { + "epoch": 25.696828629618853, + "grad_norm": 0.11752639710903168, + "learning_rate": 5.362725369142775e-06, + "loss": 0.0128, + "step": 88320 + }, + { + "epoch": 25.69973814373, + "grad_norm": 0.10498029738664627, + "learning_rate": 5.359558107426618e-06, + "loss": 0.0099, + "step": 88330 + }, + { + "epoch": 25.70264765784114, + "grad_norm": 0.07952028512954712, + "learning_rate": 5.3563915778384046e-06, + "loss": 0.0123, + "step": 88340 + }, + { + "epoch": 25.705557171952282, + "grad_norm": 0.13650281727313995, + "learning_rate": 5.353225780618612e-06, + "loss": 0.0141, + "step": 88350 + }, + { + "epoch": 25.708466686063428, + "grad_norm": 0.10780907422304153, + "learning_rate": 5.350060716007657e-06, + "loss": 0.0111, + "step": 88360 + }, + { + "epoch": 25.71137620017457, + "grad_norm": 0.0968942940235138, + "learning_rate": 5.346896384245902e-06, + "loss": 0.0115, + "step": 88370 + }, + { + "epoch": 25.714285714285715, + "grad_norm": 0.15827828645706177, + "learning_rate": 5.3437327855736615e-06, + "loss": 0.0109, + "step": 88380 + }, + { + "epoch": 25.717195228396857, + "grad_norm": 0.10371927917003632, + "learning_rate": 5.340569920231196e-06, + "loss": 0.0123, + "step": 88390 + }, + { + "epoch": 25.720104742508003, + "grad_norm": 0.1169683113694191, + "learning_rate": 5.337407788458694e-06, + "loss": 0.0139, + "step": 88400 + }, + { + "epoch": 25.723014256619145, + "grad_norm": 0.10406654328107834, + "learning_rate": 5.334246390496302e-06, + "loss": 0.0109, + "step": 88410 + }, + { + "epoch": 25.725923770730287, + "grad_norm": 0.0923481360077858, + "learning_rate": 5.331085726584114e-06, + "loss": 0.0115, + "step": 88420 + }, + { + "epoch": 25.728833284841432, + "grad_norm": 0.14993013441562653, + "learning_rate": 5.327925796962155e-06, + "loss": 0.0123, + "step": 88430 + }, + { + "epoch": 25.731742798952574, + "grad_norm": 0.13100889325141907, + "learning_rate": 5.324766601870396e-06, + "loss": 0.0129, + "step": 88440 + }, + { + "epoch": 25.73465231306372, + "grad_norm": 0.09301172941923141, + "learning_rate": 5.321608141548763e-06, + "loss": 0.012, + "step": 88450 + }, + { + "epoch": 25.73756182717486, + "grad_norm": 0.09636744856834412, + "learning_rate": 5.318450416237125e-06, + "loss": 0.0116, + "step": 88460 + }, + { + "epoch": 25.740471341286007, + "grad_norm": 0.07884202152490616, + "learning_rate": 5.315293426175287e-06, + "loss": 0.0117, + "step": 88470 + }, + { + "epoch": 25.74338085539715, + "grad_norm": 0.08410536497831345, + "learning_rate": 5.312137171602987e-06, + "loss": 0.0118, + "step": 88480 + }, + { + "epoch": 25.74629036950829, + "grad_norm": 0.08902568370103836, + "learning_rate": 5.308981652759946e-06, + "loss": 0.0132, + "step": 88490 + }, + { + "epoch": 25.749199883619436, + "grad_norm": 0.14117811620235443, + "learning_rate": 5.305826869885792e-06, + "loss": 0.0102, + "step": 88500 + }, + { + "epoch": 25.752109397730578, + "grad_norm": 0.09212297946214676, + "learning_rate": 5.302672823220107e-06, + "loss": 0.0112, + "step": 88510 + }, + { + "epoch": 25.755018911841724, + "grad_norm": 0.10552917420864105, + "learning_rate": 5.299519513002425e-06, + "loss": 0.0118, + "step": 88520 + }, + { + "epoch": 25.757928425952866, + "grad_norm": 0.0877046287059784, + "learning_rate": 5.296366939472222e-06, + "loss": 0.0121, + "step": 88530 + }, + { + "epoch": 25.760837940064008, + "grad_norm": 0.11854823678731918, + "learning_rate": 5.293215102868914e-06, + "loss": 0.0134, + "step": 88540 + }, + { + "epoch": 25.763747454175153, + "grad_norm": 0.0997321754693985, + "learning_rate": 5.290064003431852e-06, + "loss": 0.0121, + "step": 88550 + }, + { + "epoch": 25.766656968286295, + "grad_norm": 0.10242005437612534, + "learning_rate": 5.286913641400351e-06, + "loss": 0.0115, + "step": 88560 + }, + { + "epoch": 25.76956648239744, + "grad_norm": 0.07567281275987625, + "learning_rate": 5.2837640170136605e-06, + "loss": 0.0117, + "step": 88570 + }, + { + "epoch": 25.772475996508582, + "grad_norm": 0.10252887010574341, + "learning_rate": 5.28061513051097e-06, + "loss": 0.0125, + "step": 88580 + }, + { + "epoch": 25.775385510619728, + "grad_norm": 0.07606921344995499, + "learning_rate": 5.277466982131414e-06, + "loss": 0.0102, + "step": 88590 + }, + { + "epoch": 25.77829502473087, + "grad_norm": 0.22677850723266602, + "learning_rate": 5.274319572114085e-06, + "loss": 0.0131, + "step": 88600 + }, + { + "epoch": 25.78120453884201, + "grad_norm": 0.10978946834802628, + "learning_rate": 5.271172900697997e-06, + "loss": 0.0117, + "step": 88610 + }, + { + "epoch": 25.784114052953157, + "grad_norm": 0.08862143009901047, + "learning_rate": 5.268026968122125e-06, + "loss": 0.0146, + "step": 88620 + }, + { + "epoch": 25.7870235670643, + "grad_norm": 0.10283005982637405, + "learning_rate": 5.264881774625375e-06, + "loss": 0.0115, + "step": 88630 + }, + { + "epoch": 25.789933081175445, + "grad_norm": 0.17887049913406372, + "learning_rate": 5.261737320446613e-06, + "loss": 0.0113, + "step": 88640 + }, + { + "epoch": 25.792842595286587, + "grad_norm": 0.07957135885953903, + "learning_rate": 5.2585936058246284e-06, + "loss": 0.0105, + "step": 88650 + }, + { + "epoch": 25.795752109397732, + "grad_norm": 0.10643641650676727, + "learning_rate": 5.255450630998179e-06, + "loss": 0.0109, + "step": 88660 + }, + { + "epoch": 25.798661623508874, + "grad_norm": 0.08920148015022278, + "learning_rate": 5.252308396205938e-06, + "loss": 0.0114, + "step": 88670 + }, + { + "epoch": 25.801571137620016, + "grad_norm": 0.09693536162376404, + "learning_rate": 5.249166901686548e-06, + "loss": 0.0111, + "step": 88680 + }, + { + "epoch": 25.80448065173116, + "grad_norm": 0.1283738762140274, + "learning_rate": 5.2460261476785855e-06, + "loss": 0.0123, + "step": 88690 + }, + { + "epoch": 25.807390165842303, + "grad_norm": 0.08173682540655136, + "learning_rate": 5.242886134420562e-06, + "loss": 0.0118, + "step": 88700 + }, + { + "epoch": 25.81029967995345, + "grad_norm": 0.10182774811983109, + "learning_rate": 5.2397468621509455e-06, + "loss": 0.0114, + "step": 88710 + }, + { + "epoch": 25.81320919406459, + "grad_norm": 0.10274915397167206, + "learning_rate": 5.236608331108147e-06, + "loss": 0.0128, + "step": 88720 + }, + { + "epoch": 25.816118708175736, + "grad_norm": 0.09294171631336212, + "learning_rate": 5.2334705415305125e-06, + "loss": 0.0131, + "step": 88730 + }, + { + "epoch": 25.819028222286878, + "grad_norm": 0.11129912734031677, + "learning_rate": 5.230333493656333e-06, + "loss": 0.0131, + "step": 88740 + }, + { + "epoch": 25.82193773639802, + "grad_norm": 0.12304433435201645, + "learning_rate": 5.22719718772385e-06, + "loss": 0.0127, + "step": 88750 + }, + { + "epoch": 25.824847250509166, + "grad_norm": 0.09368003904819489, + "learning_rate": 5.22406162397125e-06, + "loss": 0.013, + "step": 88760 + }, + { + "epoch": 25.827756764620307, + "grad_norm": 0.2238706648349762, + "learning_rate": 5.220926802636656e-06, + "loss": 0.0119, + "step": 88770 + }, + { + "epoch": 25.830666278731453, + "grad_norm": 0.11115951091051102, + "learning_rate": 5.217792723958123e-06, + "loss": 0.0121, + "step": 88780 + }, + { + "epoch": 25.833575792842595, + "grad_norm": 0.10679587721824646, + "learning_rate": 5.2146593881736865e-06, + "loss": 0.0104, + "step": 88790 + }, + { + "epoch": 25.83648530695374, + "grad_norm": 0.10187658667564392, + "learning_rate": 5.211526795521293e-06, + "loss": 0.0123, + "step": 88800 + }, + { + "epoch": 25.839394821064882, + "grad_norm": 0.10133921355009079, + "learning_rate": 5.2083949462388346e-06, + "loss": 0.0107, + "step": 88810 + }, + { + "epoch": 25.842304335176024, + "grad_norm": 0.12432372570037842, + "learning_rate": 5.205263840564161e-06, + "loss": 0.011, + "step": 88820 + }, + { + "epoch": 25.84521384928717, + "grad_norm": 0.11117817461490631, + "learning_rate": 5.202133478735065e-06, + "loss": 0.009, + "step": 88830 + }, + { + "epoch": 25.84812336339831, + "grad_norm": 0.09145641326904297, + "learning_rate": 5.199003860989271e-06, + "loss": 0.0133, + "step": 88840 + }, + { + "epoch": 25.851032877509457, + "grad_norm": 0.23598967492580414, + "learning_rate": 5.1958749875644455e-06, + "loss": 0.0133, + "step": 88850 + }, + { + "epoch": 25.8539423916206, + "grad_norm": 0.10682088136672974, + "learning_rate": 5.192746858698216e-06, + "loss": 0.0129, + "step": 88860 + }, + { + "epoch": 25.856851905731745, + "grad_norm": 0.09650710970163345, + "learning_rate": 5.189619474628144e-06, + "loss": 0.0136, + "step": 88870 + }, + { + "epoch": 25.859761419842886, + "grad_norm": 0.10322786122560501, + "learning_rate": 5.186492835591725e-06, + "loss": 0.0136, + "step": 88880 + }, + { + "epoch": 25.86267093395403, + "grad_norm": 0.114606574177742, + "learning_rate": 5.183366941826413e-06, + "loss": 0.0138, + "step": 88890 + }, + { + "epoch": 25.865580448065174, + "grad_norm": 0.12629081308841705, + "learning_rate": 5.1802417935696e-06, + "loss": 0.0103, + "step": 88900 + }, + { + "epoch": 25.868489962176316, + "grad_norm": 0.07543253898620605, + "learning_rate": 5.177117391058615e-06, + "loss": 0.011, + "step": 88910 + }, + { + "epoch": 25.87139947628746, + "grad_norm": 0.10297894477844238, + "learning_rate": 5.1739937345307435e-06, + "loss": 0.0132, + "step": 88920 + }, + { + "epoch": 25.874308990398603, + "grad_norm": 0.11241215467453003, + "learning_rate": 5.1708708242231965e-06, + "loss": 0.0134, + "step": 88930 + }, + { + "epoch": 25.87721850450975, + "grad_norm": 0.14580558240413666, + "learning_rate": 5.16774866037315e-06, + "loss": 0.0124, + "step": 88940 + }, + { + "epoch": 25.88012801862089, + "grad_norm": 0.10049839317798615, + "learning_rate": 5.164627243217698e-06, + "loss": 0.0129, + "step": 88950 + }, + { + "epoch": 25.883037532732033, + "grad_norm": 0.14725244045257568, + "learning_rate": 5.161506572993907e-06, + "loss": 0.0115, + "step": 88960 + }, + { + "epoch": 25.885947046843178, + "grad_norm": 0.08832670748233795, + "learning_rate": 5.158386649938756e-06, + "loss": 0.0123, + "step": 88970 + }, + { + "epoch": 25.88885656095432, + "grad_norm": 0.08706319332122803, + "learning_rate": 5.1552674742891906e-06, + "loss": 0.0097, + "step": 88980 + }, + { + "epoch": 25.891766075065465, + "grad_norm": 0.08671584725379944, + "learning_rate": 5.152149046282097e-06, + "loss": 0.0095, + "step": 88990 + }, + { + "epoch": 25.894675589176607, + "grad_norm": 0.11073006689548492, + "learning_rate": 5.1490313661542886e-06, + "loss": 0.0114, + "step": 89000 + }, + { + "epoch": 25.89758510328775, + "grad_norm": 0.14182555675506592, + "learning_rate": 5.14591443414254e-06, + "loss": 0.0139, + "step": 89010 + }, + { + "epoch": 25.900494617398895, + "grad_norm": 0.17721961438655853, + "learning_rate": 5.142798250483555e-06, + "loss": 0.0117, + "step": 89020 + }, + { + "epoch": 25.903404131510037, + "grad_norm": 0.08609306067228317, + "learning_rate": 5.139682815413996e-06, + "loss": 0.0131, + "step": 89030 + }, + { + "epoch": 25.906313645621182, + "grad_norm": 0.09551149606704712, + "learning_rate": 5.136568129170449e-06, + "loss": 0.0119, + "step": 89040 + }, + { + "epoch": 25.909223159732324, + "grad_norm": 0.10553261637687683, + "learning_rate": 5.1334541919894594e-06, + "loss": 0.014, + "step": 89050 + }, + { + "epoch": 25.91213267384347, + "grad_norm": 0.12800158560276031, + "learning_rate": 5.130341004107518e-06, + "loss": 0.0113, + "step": 89060 + }, + { + "epoch": 25.91504218795461, + "grad_norm": 0.0991319790482521, + "learning_rate": 5.12722856576104e-06, + "loss": 0.0147, + "step": 89070 + }, + { + "epoch": 25.917951702065753, + "grad_norm": 0.09480152279138565, + "learning_rate": 5.1241168771863895e-06, + "loss": 0.0115, + "step": 89080 + }, + { + "epoch": 25.9208612161769, + "grad_norm": 0.11939186602830887, + "learning_rate": 5.121005938619897e-06, + "loss": 0.0142, + "step": 89090 + }, + { + "epoch": 25.92377073028804, + "grad_norm": 0.08479515463113785, + "learning_rate": 5.117895750297807e-06, + "loss": 0.0133, + "step": 89100 + }, + { + "epoch": 25.926680244399186, + "grad_norm": 0.1367853432893753, + "learning_rate": 5.1147863124563145e-06, + "loss": 0.0114, + "step": 89110 + }, + { + "epoch": 25.92958975851033, + "grad_norm": 0.16290047764778137, + "learning_rate": 5.111677625331565e-06, + "loss": 0.0132, + "step": 89120 + }, + { + "epoch": 25.932499272621474, + "grad_norm": 0.0941636934876442, + "learning_rate": 5.108569689159646e-06, + "loss": 0.0135, + "step": 89130 + }, + { + "epoch": 25.935408786732616, + "grad_norm": 0.05854180455207825, + "learning_rate": 5.105462504176583e-06, + "loss": 0.0106, + "step": 89140 + }, + { + "epoch": 25.938318300843758, + "grad_norm": 0.07304222136735916, + "learning_rate": 5.102356070618338e-06, + "loss": 0.0097, + "step": 89150 + }, + { + "epoch": 25.941227814954903, + "grad_norm": 0.1427576243877411, + "learning_rate": 5.099250388720832e-06, + "loss": 0.0099, + "step": 89160 + }, + { + "epoch": 25.944137329066045, + "grad_norm": 0.10147321224212646, + "learning_rate": 5.096145458719924e-06, + "loss": 0.014, + "step": 89170 + }, + { + "epoch": 25.94704684317719, + "grad_norm": 0.09261626750230789, + "learning_rate": 5.0930412808514045e-06, + "loss": 0.0119, + "step": 89180 + }, + { + "epoch": 25.949956357288332, + "grad_norm": 0.07010843604803085, + "learning_rate": 5.089937855351019e-06, + "loss": 0.0124, + "step": 89190 + }, + { + "epoch": 25.952865871399478, + "grad_norm": 0.07714640349149704, + "learning_rate": 5.086835182454458e-06, + "loss": 0.011, + "step": 89200 + }, + { + "epoch": 25.95577538551062, + "grad_norm": 0.10557691007852554, + "learning_rate": 5.083733262397342e-06, + "loss": 0.0113, + "step": 89210 + }, + { + "epoch": 25.958684899621762, + "grad_norm": 0.10042458027601242, + "learning_rate": 5.0806320954152394e-06, + "loss": 0.0122, + "step": 89220 + }, + { + "epoch": 25.961594413732907, + "grad_norm": 0.12709765136241913, + "learning_rate": 5.077531681743667e-06, + "loss": 0.0122, + "step": 89230 + }, + { + "epoch": 25.96450392784405, + "grad_norm": 0.12762664258480072, + "learning_rate": 5.074432021618086e-06, + "loss": 0.0124, + "step": 89240 + }, + { + "epoch": 25.967413441955195, + "grad_norm": 0.0882437527179718, + "learning_rate": 5.071333115273885e-06, + "loss": 0.0123, + "step": 89250 + }, + { + "epoch": 25.970322956066337, + "grad_norm": 0.13078820705413818, + "learning_rate": 5.068234962946416e-06, + "loss": 0.0117, + "step": 89260 + }, + { + "epoch": 25.97323247017748, + "grad_norm": 0.07908745855093002, + "learning_rate": 5.065137564870952e-06, + "loss": 0.0117, + "step": 89270 + }, + { + "epoch": 25.976141984288624, + "grad_norm": 0.063503198325634, + "learning_rate": 5.062040921282725e-06, + "loss": 0.0103, + "step": 89280 + }, + { + "epoch": 25.979051498399766, + "grad_norm": 0.10239052772521973, + "learning_rate": 5.058945032416909e-06, + "loss": 0.0115, + "step": 89290 + }, + { + "epoch": 25.98196101251091, + "grad_norm": 0.10790276527404785, + "learning_rate": 5.05584989850861e-06, + "loss": 0.0107, + "step": 89300 + }, + { + "epoch": 25.984870526622053, + "grad_norm": 0.1208251416683197, + "learning_rate": 5.052755519792889e-06, + "loss": 0.0092, + "step": 89310 + }, + { + "epoch": 25.9877800407332, + "grad_norm": 0.10080737620592117, + "learning_rate": 5.049661896504736e-06, + "loss": 0.012, + "step": 89320 + }, + { + "epoch": 25.99068955484434, + "grad_norm": 0.10869285464286804, + "learning_rate": 5.0465690288791e-06, + "loss": 0.0102, + "step": 89330 + }, + { + "epoch": 25.993599068955483, + "grad_norm": 0.11024977266788483, + "learning_rate": 5.043476917150854e-06, + "loss": 0.0103, + "step": 89340 + }, + { + "epoch": 25.99650858306663, + "grad_norm": 0.07004570215940475, + "learning_rate": 5.040385561554829e-06, + "loss": 0.0104, + "step": 89350 + }, + { + "epoch": 25.99941809717777, + "grad_norm": 0.11394871771335602, + "learning_rate": 5.037294962325799e-06, + "loss": 0.0102, + "step": 89360 + }, + { + "epoch": 26.002327611288916, + "grad_norm": 0.06968764215707779, + "learning_rate": 5.034205119698466e-06, + "loss": 0.0112, + "step": 89370 + }, + { + "epoch": 26.005237125400058, + "grad_norm": 0.0978279635310173, + "learning_rate": 5.031116033907483e-06, + "loss": 0.0118, + "step": 89380 + }, + { + "epoch": 26.008146639511203, + "grad_norm": 0.08756644278764725, + "learning_rate": 5.028027705187448e-06, + "loss": 0.0112, + "step": 89390 + }, + { + "epoch": 26.011056153622345, + "grad_norm": 0.10527973622083664, + "learning_rate": 5.024940133772904e-06, + "loss": 0.0122, + "step": 89400 + }, + { + "epoch": 26.013965667733487, + "grad_norm": 0.14659753441810608, + "learning_rate": 5.0218533198983215e-06, + "loss": 0.0126, + "step": 89410 + }, + { + "epoch": 26.016875181844632, + "grad_norm": 0.12372612208127975, + "learning_rate": 5.018767263798132e-06, + "loss": 0.0124, + "step": 89420 + }, + { + "epoch": 26.019784695955774, + "grad_norm": 0.09608685970306396, + "learning_rate": 5.0156819657067e-06, + "loss": 0.0119, + "step": 89430 + }, + { + "epoch": 26.02269421006692, + "grad_norm": 0.12133630365133286, + "learning_rate": 5.012597425858334e-06, + "loss": 0.0133, + "step": 89440 + }, + { + "epoch": 26.025603724178062, + "grad_norm": 0.09444485604763031, + "learning_rate": 5.009513644487276e-06, + "loss": 0.011, + "step": 89450 + }, + { + "epoch": 26.028513238289207, + "grad_norm": 0.10371816158294678, + "learning_rate": 5.006430621827726e-06, + "loss": 0.0108, + "step": 89460 + }, + { + "epoch": 26.03142275240035, + "grad_norm": 0.07457626610994339, + "learning_rate": 5.003348358113824e-06, + "loss": 0.0097, + "step": 89470 + }, + { + "epoch": 26.03433226651149, + "grad_norm": 0.12135842442512512, + "learning_rate": 5.000266853579641e-06, + "loss": 0.0107, + "step": 89480 + }, + { + "epoch": 26.037241780622637, + "grad_norm": 0.136338010430336, + "learning_rate": 4.997186108459188e-06, + "loss": 0.0125, + "step": 89490 + }, + { + "epoch": 26.04015129473378, + "grad_norm": 0.07690443098545074, + "learning_rate": 4.994106122986448e-06, + "loss": 0.011, + "step": 89500 + }, + { + "epoch": 26.043060808844924, + "grad_norm": 0.08546865731477737, + "learning_rate": 4.991026897395314e-06, + "loss": 0.0097, + "step": 89510 + }, + { + "epoch": 26.045970322956066, + "grad_norm": 0.10333948582410812, + "learning_rate": 4.98794843191963e-06, + "loss": 0.01, + "step": 89520 + }, + { + "epoch": 26.04887983706721, + "grad_norm": 0.0968593955039978, + "learning_rate": 4.984870726793188e-06, + "loss": 0.0087, + "step": 89530 + }, + { + "epoch": 26.051789351178353, + "grad_norm": 0.09621843695640564, + "learning_rate": 4.981793782249725e-06, + "loss": 0.0156, + "step": 89540 + }, + { + "epoch": 26.054698865289495, + "grad_norm": 0.10237745940685272, + "learning_rate": 4.978717598522905e-06, + "loss": 0.011, + "step": 89550 + }, + { + "epoch": 26.05760837940064, + "grad_norm": 0.08376898616552353, + "learning_rate": 4.975642175846352e-06, + "loss": 0.0105, + "step": 89560 + }, + { + "epoch": 26.060517893511783, + "grad_norm": 0.08209507167339325, + "learning_rate": 4.97256751445362e-06, + "loss": 0.0115, + "step": 89570 + }, + { + "epoch": 26.063427407622928, + "grad_norm": 0.10036876052618027, + "learning_rate": 4.969493614578211e-06, + "loss": 0.0134, + "step": 89580 + }, + { + "epoch": 26.06633692173407, + "grad_norm": 0.12767066061496735, + "learning_rate": 4.966420476453563e-06, + "loss": 0.0103, + "step": 89590 + }, + { + "epoch": 26.069246435845216, + "grad_norm": 0.12127848714590073, + "learning_rate": 4.963348100313064e-06, + "loss": 0.013, + "step": 89600 + }, + { + "epoch": 26.072155949956358, + "grad_norm": 0.1078428328037262, + "learning_rate": 4.9602764863900445e-06, + "loss": 0.0114, + "step": 89610 + }, + { + "epoch": 26.0750654640675, + "grad_norm": 0.07218681275844574, + "learning_rate": 4.957205634917766e-06, + "loss": 0.0115, + "step": 89620 + }, + { + "epoch": 26.077974978178645, + "grad_norm": 0.08556217700242996, + "learning_rate": 4.954135546129448e-06, + "loss": 0.0124, + "step": 89630 + }, + { + "epoch": 26.080884492289787, + "grad_norm": 0.09663402289152145, + "learning_rate": 4.951066220258231e-06, + "loss": 0.011, + "step": 89640 + }, + { + "epoch": 26.083794006400932, + "grad_norm": 0.1496475636959076, + "learning_rate": 4.947997657537219e-06, + "loss": 0.0107, + "step": 89650 + }, + { + "epoch": 26.086703520512074, + "grad_norm": 0.08986788243055344, + "learning_rate": 4.944929858199452e-06, + "loss": 0.0116, + "step": 89660 + }, + { + "epoch": 26.089613034623216, + "grad_norm": 0.0975010022521019, + "learning_rate": 4.941862822477905e-06, + "loss": 0.0129, + "step": 89670 + }, + { + "epoch": 26.09252254873436, + "grad_norm": 0.1189657673239708, + "learning_rate": 4.938796550605494e-06, + "loss": 0.0105, + "step": 89680 + }, + { + "epoch": 26.095432062845504, + "grad_norm": 0.14361122250556946, + "learning_rate": 4.935731042815088e-06, + "loss": 0.0094, + "step": 89690 + }, + { + "epoch": 26.09834157695665, + "grad_norm": 0.11967582255601883, + "learning_rate": 4.932666299339494e-06, + "loss": 0.0107, + "step": 89700 + }, + { + "epoch": 26.10125109106779, + "grad_norm": 0.07836959511041641, + "learning_rate": 4.929602320411451e-06, + "loss": 0.011, + "step": 89710 + }, + { + "epoch": 26.104160605178937, + "grad_norm": 0.08054110407829285, + "learning_rate": 4.926539106263653e-06, + "loss": 0.0125, + "step": 89720 + }, + { + "epoch": 26.10707011929008, + "grad_norm": 0.17746509611606598, + "learning_rate": 4.923476657128736e-06, + "loss": 0.0117, + "step": 89730 + }, + { + "epoch": 26.10997963340122, + "grad_norm": 0.10785964876413345, + "learning_rate": 4.920414973239267e-06, + "loss": 0.0123, + "step": 89740 + }, + { + "epoch": 26.112889147512366, + "grad_norm": 0.14870581030845642, + "learning_rate": 4.917354054827757e-06, + "loss": 0.0106, + "step": 89750 + }, + { + "epoch": 26.115798661623508, + "grad_norm": 0.07345924526453018, + "learning_rate": 4.914293902126668e-06, + "loss": 0.0112, + "step": 89760 + }, + { + "epoch": 26.118708175734653, + "grad_norm": 0.08498101681470871, + "learning_rate": 4.9112345153684005e-06, + "loss": 0.0119, + "step": 89770 + }, + { + "epoch": 26.121617689845795, + "grad_norm": 0.08542249351739883, + "learning_rate": 4.908175894785293e-06, + "loss": 0.0134, + "step": 89780 + }, + { + "epoch": 26.12452720395694, + "grad_norm": 0.11035823076963425, + "learning_rate": 4.905118040609615e-06, + "loss": 0.0122, + "step": 89790 + }, + { + "epoch": 26.127436718068083, + "grad_norm": 0.10539872199296951, + "learning_rate": 4.9020609530736116e-06, + "loss": 0.011, + "step": 89800 + }, + { + "epoch": 26.130346232179225, + "grad_norm": 0.10010574758052826, + "learning_rate": 4.899004632409438e-06, + "loss": 0.0115, + "step": 89810 + }, + { + "epoch": 26.13325574629037, + "grad_norm": 0.0883035659790039, + "learning_rate": 4.895949078849196e-06, + "loss": 0.0108, + "step": 89820 + }, + { + "epoch": 26.136165260401512, + "grad_norm": 0.08121582120656967, + "learning_rate": 4.892894292624943e-06, + "loss": 0.0114, + "step": 89830 + }, + { + "epoch": 26.139074774512657, + "grad_norm": 0.12971028685569763, + "learning_rate": 4.8898402739686695e-06, + "loss": 0.0102, + "step": 89840 + }, + { + "epoch": 26.1419842886238, + "grad_norm": 0.1204080879688263, + "learning_rate": 4.886787023112308e-06, + "loss": 0.0104, + "step": 89850 + }, + { + "epoch": 26.144893802734945, + "grad_norm": 0.1499832719564438, + "learning_rate": 4.883734540287725e-06, + "loss": 0.0103, + "step": 89860 + }, + { + "epoch": 26.147803316846087, + "grad_norm": 0.12314704805612564, + "learning_rate": 4.880682825726742e-06, + "loss": 0.0133, + "step": 89870 + }, + { + "epoch": 26.15071283095723, + "grad_norm": 0.08866723626852036, + "learning_rate": 4.877631879661123e-06, + "loss": 0.0127, + "step": 89880 + }, + { + "epoch": 26.153622345068374, + "grad_norm": 0.17901040613651276, + "learning_rate": 4.874581702322553e-06, + "loss": 0.0101, + "step": 89890 + }, + { + "epoch": 26.156531859179516, + "grad_norm": 0.09928476065397263, + "learning_rate": 4.871532293942683e-06, + "loss": 0.011, + "step": 89900 + }, + { + "epoch": 26.15944137329066, + "grad_norm": 0.11077095568180084, + "learning_rate": 4.868483654753097e-06, + "loss": 0.011, + "step": 89910 + }, + { + "epoch": 26.162350887401804, + "grad_norm": 0.11131104826927185, + "learning_rate": 4.865435784985311e-06, + "loss": 0.0122, + "step": 89920 + }, + { + "epoch": 26.16526040151295, + "grad_norm": 0.10429958254098892, + "learning_rate": 4.8623886848708e-06, + "loss": 0.0131, + "step": 89930 + }, + { + "epoch": 26.16816991562409, + "grad_norm": 0.1763681173324585, + "learning_rate": 4.859342354640961e-06, + "loss": 0.0109, + "step": 89940 + }, + { + "epoch": 26.171079429735233, + "grad_norm": 0.09277047961950302, + "learning_rate": 4.856296794527153e-06, + "loss": 0.0116, + "step": 89950 + }, + { + "epoch": 26.17398894384638, + "grad_norm": 0.1342877596616745, + "learning_rate": 4.853252004760654e-06, + "loss": 0.0123, + "step": 89960 + }, + { + "epoch": 26.17689845795752, + "grad_norm": 0.09743522107601166, + "learning_rate": 4.85020798557271e-06, + "loss": 0.0116, + "step": 89970 + }, + { + "epoch": 26.179807972068666, + "grad_norm": 0.08068060129880905, + "learning_rate": 4.84716473719448e-06, + "loss": 0.0098, + "step": 89980 + }, + { + "epoch": 26.182717486179808, + "grad_norm": 0.1007978618144989, + "learning_rate": 4.844122259857086e-06, + "loss": 0.0107, + "step": 89990 + }, + { + "epoch": 26.18562700029095, + "grad_norm": 0.0839049369096756, + "learning_rate": 4.841080553791589e-06, + "loss": 0.0104, + "step": 90000 + }, + { + "epoch": 26.188536514402095, + "grad_norm": 0.08725588023662567, + "learning_rate": 4.838039619228974e-06, + "loss": 0.011, + "step": 90010 + }, + { + "epoch": 26.191446028513237, + "grad_norm": 0.1069418266415596, + "learning_rate": 4.834999456400188e-06, + "loss": 0.0109, + "step": 90020 + }, + { + "epoch": 26.194355542624383, + "grad_norm": 0.149014413356781, + "learning_rate": 4.831960065536115e-06, + "loss": 0.0113, + "step": 90030 + }, + { + "epoch": 26.197265056735525, + "grad_norm": 0.09270782023668289, + "learning_rate": 4.828921446867572e-06, + "loss": 0.011, + "step": 90040 + }, + { + "epoch": 26.20017457084667, + "grad_norm": 0.10007612407207489, + "learning_rate": 4.825883600625316e-06, + "loss": 0.0107, + "step": 90050 + }, + { + "epoch": 26.203084084957812, + "grad_norm": 0.11658672988414764, + "learning_rate": 4.822846527040058e-06, + "loss": 0.0115, + "step": 90060 + }, + { + "epoch": 26.205993599068954, + "grad_norm": 0.16269689798355103, + "learning_rate": 4.819810226342449e-06, + "loss": 0.0103, + "step": 90070 + }, + { + "epoch": 26.2089031131801, + "grad_norm": 0.11922622472047806, + "learning_rate": 4.816774698763067e-06, + "loss": 0.0109, + "step": 90080 + }, + { + "epoch": 26.21181262729124, + "grad_norm": 0.27552881836891174, + "learning_rate": 4.813739944532436e-06, + "loss": 0.0109, + "step": 90090 + }, + { + "epoch": 26.214722141402387, + "grad_norm": 0.08755198866128922, + "learning_rate": 4.810705963881042e-06, + "loss": 0.0118, + "step": 90100 + }, + { + "epoch": 26.21763165551353, + "grad_norm": 0.1502755731344223, + "learning_rate": 4.807672757039286e-06, + "loss": 0.0137, + "step": 90110 + }, + { + "epoch": 26.220541169624674, + "grad_norm": 0.19614984095096588, + "learning_rate": 4.8046403242375155e-06, + "loss": 0.0144, + "step": 90120 + }, + { + "epoch": 26.223450683735816, + "grad_norm": 0.11615870893001556, + "learning_rate": 4.801608665706029e-06, + "loss": 0.0105, + "step": 90130 + }, + { + "epoch": 26.226360197846958, + "grad_norm": 0.10994139313697815, + "learning_rate": 4.798577781675066e-06, + "loss": 0.0129, + "step": 90140 + }, + { + "epoch": 26.229269711958104, + "grad_norm": 0.10202840715646744, + "learning_rate": 4.795547672374798e-06, + "loss": 0.0116, + "step": 90150 + }, + { + "epoch": 26.232179226069245, + "grad_norm": 0.12712840735912323, + "learning_rate": 4.792518338035334e-06, + "loss": 0.015, + "step": 90160 + }, + { + "epoch": 26.23508874018039, + "grad_norm": 0.1922171711921692, + "learning_rate": 4.78948977888674e-06, + "loss": 0.0121, + "step": 90170 + }, + { + "epoch": 26.237998254291533, + "grad_norm": 0.13594913482666016, + "learning_rate": 4.786461995159017e-06, + "loss": 0.0099, + "step": 90180 + }, + { + "epoch": 26.24090776840268, + "grad_norm": 0.09519024938344955, + "learning_rate": 4.7834349870820984e-06, + "loss": 0.0114, + "step": 90190 + }, + { + "epoch": 26.24381728251382, + "grad_norm": 0.13181206583976746, + "learning_rate": 4.780408754885869e-06, + "loss": 0.011, + "step": 90200 + }, + { + "epoch": 26.246726796624962, + "grad_norm": 0.09033214300870895, + "learning_rate": 4.777383298800157e-06, + "loss": 0.0124, + "step": 90210 + }, + { + "epoch": 26.249636310736108, + "grad_norm": 0.1790747344493866, + "learning_rate": 4.774358619054718e-06, + "loss": 0.0121, + "step": 90220 + }, + { + "epoch": 26.25254582484725, + "grad_norm": 0.12913504242897034, + "learning_rate": 4.771334715879254e-06, + "loss": 0.0128, + "step": 90230 + }, + { + "epoch": 26.255455338958395, + "grad_norm": 0.10624168068170547, + "learning_rate": 4.768311589503417e-06, + "loss": 0.0122, + "step": 90240 + }, + { + "epoch": 26.258364853069537, + "grad_norm": 0.10724175721406937, + "learning_rate": 4.765289240156794e-06, + "loss": 0.0121, + "step": 90250 + }, + { + "epoch": 26.261274367180683, + "grad_norm": 0.08217649161815643, + "learning_rate": 4.762267668068906e-06, + "loss": 0.0119, + "step": 90260 + }, + { + "epoch": 26.264183881291824, + "grad_norm": 0.10530491173267365, + "learning_rate": 4.759246873469231e-06, + "loss": 0.0119, + "step": 90270 + }, + { + "epoch": 26.267093395402966, + "grad_norm": 0.09962466359138489, + "learning_rate": 4.756226856587168e-06, + "loss": 0.0115, + "step": 90280 + }, + { + "epoch": 26.270002909514112, + "grad_norm": 0.08870021253824234, + "learning_rate": 4.753207617652072e-06, + "loss": 0.0123, + "step": 90290 + }, + { + "epoch": 26.272912423625254, + "grad_norm": 0.10023543238639832, + "learning_rate": 4.7501891568932385e-06, + "loss": 0.0108, + "step": 90300 + }, + { + "epoch": 26.2758219377364, + "grad_norm": 0.14649999141693115, + "learning_rate": 4.747171474539891e-06, + "loss": 0.0106, + "step": 90310 + }, + { + "epoch": 26.27873145184754, + "grad_norm": 0.2742539644241333, + "learning_rate": 4.744154570821213e-06, + "loss": 0.0122, + "step": 90320 + }, + { + "epoch": 26.281640965958683, + "grad_norm": 0.1050809994339943, + "learning_rate": 4.741138445966307e-06, + "loss": 0.0123, + "step": 90330 + }, + { + "epoch": 26.28455048006983, + "grad_norm": 0.13148674368858337, + "learning_rate": 4.738123100204238e-06, + "loss": 0.013, + "step": 90340 + }, + { + "epoch": 26.28745999418097, + "grad_norm": 0.5291737914085388, + "learning_rate": 4.7351085337639925e-06, + "loss": 0.0119, + "step": 90350 + }, + { + "epoch": 26.290369508292116, + "grad_norm": 0.15438531339168549, + "learning_rate": 4.732094746874513e-06, + "loss": 0.0117, + "step": 90360 + }, + { + "epoch": 26.293279022403258, + "grad_norm": 0.08039068430662155, + "learning_rate": 4.729081739764678e-06, + "loss": 0.0104, + "step": 90370 + }, + { + "epoch": 26.296188536514403, + "grad_norm": 0.0983627662062645, + "learning_rate": 4.726069512663303e-06, + "loss": 0.0142, + "step": 90380 + }, + { + "epoch": 26.299098050625545, + "grad_norm": 0.08594179898500443, + "learning_rate": 4.723058065799139e-06, + "loss": 0.0114, + "step": 90390 + }, + { + "epoch": 26.302007564736687, + "grad_norm": 0.09103954583406448, + "learning_rate": 4.720047399400899e-06, + "loss": 0.0107, + "step": 90400 + }, + { + "epoch": 26.304917078847833, + "grad_norm": 0.11164502799510956, + "learning_rate": 4.717037513697221e-06, + "loss": 0.0124, + "step": 90410 + }, + { + "epoch": 26.307826592958975, + "grad_norm": 0.12137524783611298, + "learning_rate": 4.714028408916675e-06, + "loss": 0.0139, + "step": 90420 + }, + { + "epoch": 26.31073610707012, + "grad_norm": 0.10045269876718521, + "learning_rate": 4.711020085287791e-06, + "loss": 0.0114, + "step": 90430 + }, + { + "epoch": 26.313645621181262, + "grad_norm": 0.1035187840461731, + "learning_rate": 4.708012543039034e-06, + "loss": 0.0098, + "step": 90440 + }, + { + "epoch": 26.316555135292408, + "grad_norm": 0.0985151007771492, + "learning_rate": 4.705005782398803e-06, + "loss": 0.0115, + "step": 90450 + }, + { + "epoch": 26.31946464940355, + "grad_norm": 0.12011783570051193, + "learning_rate": 4.701999803595437e-06, + "loss": 0.0134, + "step": 90460 + }, + { + "epoch": 26.32237416351469, + "grad_norm": 0.10702847689390182, + "learning_rate": 4.698994606857225e-06, + "loss": 0.0142, + "step": 90470 + }, + { + "epoch": 26.325283677625837, + "grad_norm": 0.09349724650382996, + "learning_rate": 4.695990192412397e-06, + "loss": 0.0136, + "step": 90480 + }, + { + "epoch": 26.32819319173698, + "grad_norm": 0.11306893825531006, + "learning_rate": 4.692986560489108e-06, + "loss": 0.0097, + "step": 90490 + }, + { + "epoch": 26.331102705848124, + "grad_norm": 0.13912679255008698, + "learning_rate": 4.689983711315472e-06, + "loss": 0.0098, + "step": 90500 + }, + { + "epoch": 26.334012219959266, + "grad_norm": 0.1072721928358078, + "learning_rate": 4.6869816451195346e-06, + "loss": 0.0118, + "step": 90510 + }, + { + "epoch": 26.336921734070412, + "grad_norm": 0.07842212170362473, + "learning_rate": 4.683980362129283e-06, + "loss": 0.0105, + "step": 90520 + }, + { + "epoch": 26.339831248181554, + "grad_norm": 0.08809266239404678, + "learning_rate": 4.680979862572639e-06, + "loss": 0.0111, + "step": 90530 + }, + { + "epoch": 26.342740762292696, + "grad_norm": 0.14273081719875336, + "learning_rate": 4.677980146677472e-06, + "loss": 0.0134, + "step": 90540 + }, + { + "epoch": 26.34565027640384, + "grad_norm": 0.11116701364517212, + "learning_rate": 4.6749812146716005e-06, + "loss": 0.0139, + "step": 90550 + }, + { + "epoch": 26.348559790514983, + "grad_norm": 0.09358370304107666, + "learning_rate": 4.6719830667827615e-06, + "loss": 0.0138, + "step": 90560 + }, + { + "epoch": 26.35146930462613, + "grad_norm": 0.1163971871137619, + "learning_rate": 4.668985703238654e-06, + "loss": 0.011, + "step": 90570 + }, + { + "epoch": 26.35437881873727, + "grad_norm": 0.09496762603521347, + "learning_rate": 4.665989124266898e-06, + "loss": 0.0093, + "step": 90580 + }, + { + "epoch": 26.357288332848416, + "grad_norm": 0.1250516027212143, + "learning_rate": 4.662993330095076e-06, + "loss": 0.0122, + "step": 90590 + }, + { + "epoch": 26.360197846959558, + "grad_norm": 0.1159338429570198, + "learning_rate": 4.659998320950685e-06, + "loss": 0.0122, + "step": 90600 + }, + { + "epoch": 26.3631073610707, + "grad_norm": 0.12476833909749985, + "learning_rate": 4.657004097061186e-06, + "loss": 0.0122, + "step": 90610 + }, + { + "epoch": 26.366016875181845, + "grad_norm": 0.10276801139116287, + "learning_rate": 4.6540106586539714e-06, + "loss": 0.012, + "step": 90620 + }, + { + "epoch": 26.368926389292987, + "grad_norm": 0.11106456816196442, + "learning_rate": 4.6510180059563654e-06, + "loss": 0.0117, + "step": 90630 + }, + { + "epoch": 26.371835903404133, + "grad_norm": 0.09830178320407867, + "learning_rate": 4.648026139195649e-06, + "loss": 0.0126, + "step": 90640 + }, + { + "epoch": 26.374745417515275, + "grad_norm": 0.09898657351732254, + "learning_rate": 4.645035058599025e-06, + "loss": 0.0103, + "step": 90650 + }, + { + "epoch": 26.377654931626417, + "grad_norm": 0.09150796383619308, + "learning_rate": 4.6420447643936534e-06, + "loss": 0.0134, + "step": 90660 + }, + { + "epoch": 26.380564445737562, + "grad_norm": 0.11796854436397552, + "learning_rate": 4.639055256806629e-06, + "loss": 0.0113, + "step": 90670 + }, + { + "epoch": 26.383473959848704, + "grad_norm": 0.17374378442764282, + "learning_rate": 4.63606653606498e-06, + "loss": 0.0111, + "step": 90680 + }, + { + "epoch": 26.38638347395985, + "grad_norm": 0.12487521022558212, + "learning_rate": 4.6330786023956786e-06, + "loss": 0.013, + "step": 90690 + }, + { + "epoch": 26.38929298807099, + "grad_norm": 0.12524235248565674, + "learning_rate": 4.630091456025641e-06, + "loss": 0.0119, + "step": 90700 + }, + { + "epoch": 26.392202502182137, + "grad_norm": 0.09337666630744934, + "learning_rate": 4.627105097181728e-06, + "loss": 0.0096, + "step": 90710 + }, + { + "epoch": 26.39511201629328, + "grad_norm": 0.10022281110286713, + "learning_rate": 4.624119526090722e-06, + "loss": 0.0117, + "step": 90720 + }, + { + "epoch": 26.39802153040442, + "grad_norm": 0.11115587502717972, + "learning_rate": 4.621134742979361e-06, + "loss": 0.0119, + "step": 90730 + }, + { + "epoch": 26.400931044515566, + "grad_norm": 0.1378472000360489, + "learning_rate": 4.618150748074329e-06, + "loss": 0.0122, + "step": 90740 + }, + { + "epoch": 26.403840558626708, + "grad_norm": 0.10111185908317566, + "learning_rate": 4.615167541602233e-06, + "loss": 0.0114, + "step": 90750 + }, + { + "epoch": 26.406750072737854, + "grad_norm": 0.10256028920412064, + "learning_rate": 4.6121851237896244e-06, + "loss": 0.0114, + "step": 90760 + }, + { + "epoch": 26.409659586848996, + "grad_norm": 0.08003424108028412, + "learning_rate": 4.609203494863e-06, + "loss": 0.0096, + "step": 90770 + }, + { + "epoch": 26.41256910096014, + "grad_norm": 0.1134411096572876, + "learning_rate": 4.606222655048804e-06, + "loss": 0.0114, + "step": 90780 + }, + { + "epoch": 26.415478615071283, + "grad_norm": 0.08122224360704422, + "learning_rate": 4.603242604573403e-06, + "loss": 0.0111, + "step": 90790 + }, + { + "epoch": 26.418388129182425, + "grad_norm": 0.091563381254673, + "learning_rate": 4.600263343663104e-06, + "loss": 0.0103, + "step": 90800 + }, + { + "epoch": 26.42129764329357, + "grad_norm": 0.12655572593212128, + "learning_rate": 4.597284872544181e-06, + "loss": 0.0106, + "step": 90810 + }, + { + "epoch": 26.424207157404712, + "grad_norm": 0.0899205282330513, + "learning_rate": 4.59430719144282e-06, + "loss": 0.009, + "step": 90820 + }, + { + "epoch": 26.427116671515858, + "grad_norm": 0.08645036071538925, + "learning_rate": 4.591330300585152e-06, + "loss": 0.0133, + "step": 90830 + }, + { + "epoch": 26.430026185627, + "grad_norm": 0.0885079950094223, + "learning_rate": 4.588354200197257e-06, + "loss": 0.0105, + "step": 90840 + }, + { + "epoch": 26.432935699738145, + "grad_norm": 0.3646257817745209, + "learning_rate": 4.585378890505154e-06, + "loss": 0.0097, + "step": 90850 + }, + { + "epoch": 26.435845213849287, + "grad_norm": 0.10318149626255035, + "learning_rate": 4.582404371734787e-06, + "loss": 0.0122, + "step": 90860 + }, + { + "epoch": 26.43875472796043, + "grad_norm": 0.08863361924886703, + "learning_rate": 4.5794306441120645e-06, + "loss": 0.011, + "step": 90870 + }, + { + "epoch": 26.441664242071575, + "grad_norm": 0.09441103786230087, + "learning_rate": 4.57645770786281e-06, + "loss": 0.0113, + "step": 90880 + }, + { + "epoch": 26.444573756182717, + "grad_norm": 0.1265801042318344, + "learning_rate": 4.573485563212807e-06, + "loss": 0.0115, + "step": 90890 + }, + { + "epoch": 26.447483270293862, + "grad_norm": 0.10067295283079147, + "learning_rate": 4.5705142103877626e-06, + "loss": 0.0109, + "step": 90900 + }, + { + "epoch": 26.450392784405004, + "grad_norm": 0.18609189987182617, + "learning_rate": 4.567543649613336e-06, + "loss": 0.0117, + "step": 90910 + }, + { + "epoch": 26.45330229851615, + "grad_norm": 0.10647151619195938, + "learning_rate": 4.564573881115124e-06, + "loss": 0.0108, + "step": 90920 + }, + { + "epoch": 26.45621181262729, + "grad_norm": 0.08267899602651596, + "learning_rate": 4.561604905118654e-06, + "loss": 0.0116, + "step": 90930 + }, + { + "epoch": 26.459121326738433, + "grad_norm": 0.09946750104427338, + "learning_rate": 4.558636721849409e-06, + "loss": 0.0132, + "step": 90940 + }, + { + "epoch": 26.46203084084958, + "grad_norm": 0.10589633136987686, + "learning_rate": 4.555669331532793e-06, + "loss": 0.0142, + "step": 90950 + }, + { + "epoch": 26.46494035496072, + "grad_norm": 0.10588102787733078, + "learning_rate": 4.5527027343941655e-06, + "loss": 0.0135, + "step": 90960 + }, + { + "epoch": 26.467849869071866, + "grad_norm": 0.22374524176120758, + "learning_rate": 4.549736930658823e-06, + "loss": 0.0112, + "step": 90970 + }, + { + "epoch": 26.470759383183008, + "grad_norm": 0.12799575924873352, + "learning_rate": 4.5467719205519945e-06, + "loss": 0.0129, + "step": 90980 + }, + { + "epoch": 26.473668897294154, + "grad_norm": 0.1023198589682579, + "learning_rate": 4.543807704298849e-06, + "loss": 0.013, + "step": 90990 + }, + { + "epoch": 26.476578411405296, + "grad_norm": 0.1371038556098938, + "learning_rate": 4.540844282124504e-06, + "loss": 0.0115, + "step": 91000 + }, + { + "epoch": 26.479487925516437, + "grad_norm": 0.0804385393857956, + "learning_rate": 4.537881654254015e-06, + "loss": 0.0125, + "step": 91010 + }, + { + "epoch": 26.482397439627583, + "grad_norm": 0.0879087895154953, + "learning_rate": 4.534919820912366e-06, + "loss": 0.0102, + "step": 91020 + }, + { + "epoch": 26.485306953738725, + "grad_norm": 0.12771983444690704, + "learning_rate": 4.531958782324495e-06, + "loss": 0.0134, + "step": 91030 + }, + { + "epoch": 26.48821646784987, + "grad_norm": 0.09454658627510071, + "learning_rate": 4.528998538715274e-06, + "loss": 0.0108, + "step": 91040 + }, + { + "epoch": 26.491125981961012, + "grad_norm": 0.07814504206180573, + "learning_rate": 4.526039090309512e-06, + "loss": 0.0109, + "step": 91050 + }, + { + "epoch": 26.494035496072154, + "grad_norm": 0.1336139738559723, + "learning_rate": 4.523080437331953e-06, + "loss": 0.0133, + "step": 91060 + }, + { + "epoch": 26.4969450101833, + "grad_norm": 0.16186927258968353, + "learning_rate": 4.520122580007295e-06, + "loss": 0.0134, + "step": 91070 + }, + { + "epoch": 26.49985452429444, + "grad_norm": 0.13468362390995026, + "learning_rate": 4.51716551856017e-06, + "loss": 0.0122, + "step": 91080 + }, + { + "epoch": 26.502764038405587, + "grad_norm": 0.08516331762075424, + "learning_rate": 4.514209253215143e-06, + "loss": 0.0102, + "step": 91090 + }, + { + "epoch": 26.50567355251673, + "grad_norm": 0.1115460991859436, + "learning_rate": 4.511253784196714e-06, + "loss": 0.0126, + "step": 91100 + }, + { + "epoch": 26.508583066627875, + "grad_norm": 0.10566014796495438, + "learning_rate": 4.50829911172935e-06, + "loss": 0.0092, + "step": 91110 + }, + { + "epoch": 26.511492580739016, + "grad_norm": 0.09678252786397934, + "learning_rate": 4.505345236037428e-06, + "loss": 0.0092, + "step": 91120 + }, + { + "epoch": 26.51440209485016, + "grad_norm": 0.12016180902719498, + "learning_rate": 4.502392157345272e-06, + "loss": 0.0107, + "step": 91130 + }, + { + "epoch": 26.517311608961304, + "grad_norm": 0.12916220724582672, + "learning_rate": 4.4994398758771555e-06, + "loss": 0.0106, + "step": 91140 + }, + { + "epoch": 26.520221123072446, + "grad_norm": 0.09976735711097717, + "learning_rate": 4.496488391857285e-06, + "loss": 0.0134, + "step": 91150 + }, + { + "epoch": 26.52313063718359, + "grad_norm": 0.11443876475095749, + "learning_rate": 4.4935377055098055e-06, + "loss": 0.0111, + "step": 91160 + }, + { + "epoch": 26.526040151294733, + "grad_norm": 0.13520805537700653, + "learning_rate": 4.490587817058795e-06, + "loss": 0.0104, + "step": 91170 + }, + { + "epoch": 26.52894966540588, + "grad_norm": 0.1370030790567398, + "learning_rate": 4.487638726728285e-06, + "loss": 0.0131, + "step": 91180 + }, + { + "epoch": 26.53185917951702, + "grad_norm": 0.09787514805793762, + "learning_rate": 4.484690434742242e-06, + "loss": 0.0104, + "step": 91190 + }, + { + "epoch": 26.534768693628163, + "grad_norm": 0.08659069240093231, + "learning_rate": 4.481742941324561e-06, + "loss": 0.0098, + "step": 91200 + }, + { + "epoch": 26.537678207739308, + "grad_norm": 0.1023460105061531, + "learning_rate": 4.478796246699088e-06, + "loss": 0.0113, + "step": 91210 + }, + { + "epoch": 26.54058772185045, + "grad_norm": 0.07952608168125153, + "learning_rate": 4.475850351089613e-06, + "loss": 0.0092, + "step": 91220 + }, + { + "epoch": 26.543497235961595, + "grad_norm": 0.2599896192550659, + "learning_rate": 4.472905254719844e-06, + "loss": 0.0107, + "step": 91230 + }, + { + "epoch": 26.546406750072737, + "grad_norm": 0.11218587309122086, + "learning_rate": 4.469960957813453e-06, + "loss": 0.0117, + "step": 91240 + }, + { + "epoch": 26.549316264183883, + "grad_norm": 0.14861713349819183, + "learning_rate": 4.467017460594032e-06, + "loss": 0.0117, + "step": 91250 + }, + { + "epoch": 26.552225778295025, + "grad_norm": 0.17651528120040894, + "learning_rate": 4.464074763285126e-06, + "loss": 0.0102, + "step": 91260 + }, + { + "epoch": 26.555135292406167, + "grad_norm": 0.09837374836206436, + "learning_rate": 4.461132866110207e-06, + "loss": 0.0094, + "step": 91270 + }, + { + "epoch": 26.558044806517312, + "grad_norm": 0.09863891452550888, + "learning_rate": 4.458191769292701e-06, + "loss": 0.0119, + "step": 91280 + }, + { + "epoch": 26.560954320628454, + "grad_norm": 0.12719625234603882, + "learning_rate": 4.455251473055955e-06, + "loss": 0.011, + "step": 91290 + }, + { + "epoch": 26.5638638347396, + "grad_norm": 0.1400635540485382, + "learning_rate": 4.452311977623271e-06, + "loss": 0.0129, + "step": 91300 + }, + { + "epoch": 26.56677334885074, + "grad_norm": 0.1048617735505104, + "learning_rate": 4.449373283217887e-06, + "loss": 0.0109, + "step": 91310 + }, + { + "epoch": 26.569682862961887, + "grad_norm": 0.10718239098787308, + "learning_rate": 4.446435390062971e-06, + "loss": 0.0103, + "step": 91320 + }, + { + "epoch": 26.57259237707303, + "grad_norm": 0.11520854383707047, + "learning_rate": 4.4434982983816394e-06, + "loss": 0.0125, + "step": 91330 + }, + { + "epoch": 26.57550189118417, + "grad_norm": 0.11858239024877548, + "learning_rate": 4.44056200839695e-06, + "loss": 0.0101, + "step": 91340 + }, + { + "epoch": 26.578411405295316, + "grad_norm": 0.09878456592559814, + "learning_rate": 4.437626520331889e-06, + "loss": 0.0111, + "step": 91350 + }, + { + "epoch": 26.58132091940646, + "grad_norm": 0.12802881002426147, + "learning_rate": 4.434691834409384e-06, + "loss": 0.0115, + "step": 91360 + }, + { + "epoch": 26.584230433517604, + "grad_norm": 0.07962270826101303, + "learning_rate": 4.431757950852308e-06, + "loss": 0.0099, + "step": 91370 + }, + { + "epoch": 26.587139947628746, + "grad_norm": 0.08883122354745865, + "learning_rate": 4.428824869883478e-06, + "loss": 0.0129, + "step": 91380 + }, + { + "epoch": 26.590049461739888, + "grad_norm": 0.08238878101110458, + "learning_rate": 4.425892591725635e-06, + "loss": 0.0111, + "step": 91390 + }, + { + "epoch": 26.592958975851033, + "grad_norm": 0.11293130367994308, + "learning_rate": 4.422961116601457e-06, + "loss": 0.0113, + "step": 91400 + }, + { + "epoch": 26.595868489962175, + "grad_norm": 0.08979939669370651, + "learning_rate": 4.420030444733587e-06, + "loss": 0.0124, + "step": 91410 + }, + { + "epoch": 26.59877800407332, + "grad_norm": 0.12223154306411743, + "learning_rate": 4.417100576344586e-06, + "loss": 0.011, + "step": 91420 + }, + { + "epoch": 26.601687518184463, + "grad_norm": 0.09921611845493317, + "learning_rate": 4.414171511656948e-06, + "loss": 0.0114, + "step": 91430 + }, + { + "epoch": 26.604597032295608, + "grad_norm": 0.1262313723564148, + "learning_rate": 4.411243250893125e-06, + "loss": 0.0154, + "step": 91440 + }, + { + "epoch": 26.60750654640675, + "grad_norm": 0.09721556305885315, + "learning_rate": 4.408315794275502e-06, + "loss": 0.0106, + "step": 91450 + }, + { + "epoch": 26.610416060517892, + "grad_norm": 0.09846573323011398, + "learning_rate": 4.405389142026395e-06, + "loss": 0.0112, + "step": 91460 + }, + { + "epoch": 26.613325574629037, + "grad_norm": 0.11640197038650513, + "learning_rate": 4.402463294368056e-06, + "loss": 0.0124, + "step": 91470 + }, + { + "epoch": 26.61623508874018, + "grad_norm": 0.14038334786891937, + "learning_rate": 4.399538251522701e-06, + "loss": 0.01, + "step": 91480 + }, + { + "epoch": 26.619144602851325, + "grad_norm": 0.08407562226057053, + "learning_rate": 4.396614013712458e-06, + "loss": 0.0124, + "step": 91490 + }, + { + "epoch": 26.622054116962467, + "grad_norm": 0.08696401119232178, + "learning_rate": 4.393690581159399e-06, + "loss": 0.0118, + "step": 91500 + }, + { + "epoch": 26.624963631073612, + "grad_norm": 0.10573653131723404, + "learning_rate": 4.390767954085548e-06, + "loss": 0.0128, + "step": 91510 + }, + { + "epoch": 26.627873145184754, + "grad_norm": 0.09253575652837753, + "learning_rate": 4.3878461327128585e-06, + "loss": 0.0129, + "step": 91520 + }, + { + "epoch": 26.630782659295896, + "grad_norm": 0.09447692334651947, + "learning_rate": 4.38492511726322e-06, + "loss": 0.0114, + "step": 91530 + }, + { + "epoch": 26.63369217340704, + "grad_norm": 0.12340085208415985, + "learning_rate": 4.382004907958462e-06, + "loss": 0.0114, + "step": 91540 + }, + { + "epoch": 26.636601687518183, + "grad_norm": 0.1250937134027481, + "learning_rate": 4.37908550502036e-06, + "loss": 0.0101, + "step": 91550 + }, + { + "epoch": 26.63951120162933, + "grad_norm": 0.10686202347278595, + "learning_rate": 4.376166908670624e-06, + "loss": 0.0124, + "step": 91560 + }, + { + "epoch": 26.64242071574047, + "grad_norm": 0.13538463413715363, + "learning_rate": 4.373249119130897e-06, + "loss": 0.0106, + "step": 91570 + }, + { + "epoch": 26.645330229851616, + "grad_norm": 0.1337747424840927, + "learning_rate": 4.370332136622768e-06, + "loss": 0.0111, + "step": 91580 + }, + { + "epoch": 26.64823974396276, + "grad_norm": 0.09373465925455093, + "learning_rate": 4.367415961367769e-06, + "loss": 0.0113, + "step": 91590 + }, + { + "epoch": 26.6511492580739, + "grad_norm": 0.09710672497749329, + "learning_rate": 4.364500593587355e-06, + "loss": 0.0114, + "step": 91600 + }, + { + "epoch": 26.654058772185046, + "grad_norm": 0.12030233442783356, + "learning_rate": 4.361586033502936e-06, + "loss": 0.0103, + "step": 91610 + }, + { + "epoch": 26.656968286296188, + "grad_norm": 0.08214841037988663, + "learning_rate": 4.358672281335845e-06, + "loss": 0.0108, + "step": 91620 + }, + { + "epoch": 26.659877800407333, + "grad_norm": 0.08227057009935379, + "learning_rate": 4.355759337307374e-06, + "loss": 0.0104, + "step": 91630 + }, + { + "epoch": 26.662787314518475, + "grad_norm": 0.0918714627623558, + "learning_rate": 4.35284720163873e-06, + "loss": 0.0114, + "step": 91640 + }, + { + "epoch": 26.66569682862962, + "grad_norm": 0.16049595177173615, + "learning_rate": 4.34993587455108e-06, + "loss": 0.0133, + "step": 91650 + }, + { + "epoch": 26.668606342740762, + "grad_norm": 0.09874666482210159, + "learning_rate": 4.347025356265514e-06, + "loss": 0.0122, + "step": 91660 + }, + { + "epoch": 26.671515856851904, + "grad_norm": 0.08409792184829712, + "learning_rate": 4.3441156470030656e-06, + "loss": 0.0121, + "step": 91670 + }, + { + "epoch": 26.67442537096305, + "grad_norm": 0.09902283549308777, + "learning_rate": 4.341206746984717e-06, + "loss": 0.0112, + "step": 91680 + }, + { + "epoch": 26.677334885074192, + "grad_norm": 0.10682518780231476, + "learning_rate": 4.338298656431369e-06, + "loss": 0.0111, + "step": 91690 + }, + { + "epoch": 26.680244399185337, + "grad_norm": 0.09408320486545563, + "learning_rate": 4.335391375563876e-06, + "loss": 0.0108, + "step": 91700 + }, + { + "epoch": 26.68315391329648, + "grad_norm": 0.19932375848293304, + "learning_rate": 4.332484904603033e-06, + "loss": 0.0115, + "step": 91710 + }, + { + "epoch": 26.686063427407625, + "grad_norm": 0.2168942242860794, + "learning_rate": 4.329579243769564e-06, + "loss": 0.0116, + "step": 91720 + }, + { + "epoch": 26.688972941518767, + "grad_norm": 0.12121767550706863, + "learning_rate": 4.326674393284127e-06, + "loss": 0.0123, + "step": 91730 + }, + { + "epoch": 26.69188245562991, + "grad_norm": 0.13162629306316376, + "learning_rate": 4.3237703533673314e-06, + "loss": 0.0116, + "step": 91740 + }, + { + "epoch": 26.694791969741054, + "grad_norm": 0.13065341114997864, + "learning_rate": 4.320867124239726e-06, + "loss": 0.0111, + "step": 91750 + }, + { + "epoch": 26.697701483852196, + "grad_norm": 0.15767158567905426, + "learning_rate": 4.317964706121786e-06, + "loss": 0.0104, + "step": 91760 + }, + { + "epoch": 26.70061099796334, + "grad_norm": 0.12849470973014832, + "learning_rate": 4.3150630992339205e-06, + "loss": 0.0108, + "step": 91770 + }, + { + "epoch": 26.703520512074483, + "grad_norm": 0.11732284724712372, + "learning_rate": 4.312162303796509e-06, + "loss": 0.0119, + "step": 91780 + }, + { + "epoch": 26.706430026185625, + "grad_norm": 0.13909600675106049, + "learning_rate": 4.309262320029837e-06, + "loss": 0.0135, + "step": 91790 + }, + { + "epoch": 26.70933954029677, + "grad_norm": 0.1262485831975937, + "learning_rate": 4.3063631481541334e-06, + "loss": 0.0107, + "step": 91800 + }, + { + "epoch": 26.712249054407913, + "grad_norm": 0.09967502951622009, + "learning_rate": 4.3034647883895785e-06, + "loss": 0.0115, + "step": 91810 + }, + { + "epoch": 26.71515856851906, + "grad_norm": 0.0806422233581543, + "learning_rate": 4.3005672409562855e-06, + "loss": 0.0129, + "step": 91820 + }, + { + "epoch": 26.7180680826302, + "grad_norm": 0.11641861498355865, + "learning_rate": 4.297670506074299e-06, + "loss": 0.012, + "step": 91830 + }, + { + "epoch": 26.720977596741346, + "grad_norm": 0.10879762470722198, + "learning_rate": 4.2947745839636065e-06, + "loss": 0.0107, + "step": 91840 + }, + { + "epoch": 26.723887110852488, + "grad_norm": 0.12316453456878662, + "learning_rate": 4.291879474844135e-06, + "loss": 0.0101, + "step": 91850 + }, + { + "epoch": 26.72679662496363, + "grad_norm": 0.12228534370660782, + "learning_rate": 4.288985178935753e-06, + "loss": 0.0125, + "step": 91860 + }, + { + "epoch": 26.729706139074775, + "grad_norm": 0.06362485140562057, + "learning_rate": 4.286091696458257e-06, + "loss": 0.0098, + "step": 91870 + }, + { + "epoch": 26.732615653185917, + "grad_norm": 0.1321079283952713, + "learning_rate": 4.283199027631392e-06, + "loss": 0.0131, + "step": 91880 + }, + { + "epoch": 26.735525167297062, + "grad_norm": 0.10637672245502472, + "learning_rate": 4.280307172674839e-06, + "loss": 0.0114, + "step": 91890 + }, + { + "epoch": 26.738434681408204, + "grad_norm": 0.08570706844329834, + "learning_rate": 4.277416131808212e-06, + "loss": 0.0117, + "step": 91900 + }, + { + "epoch": 26.74134419551935, + "grad_norm": 0.10658394545316696, + "learning_rate": 4.2745259052510645e-06, + "loss": 0.0107, + "step": 91910 + }, + { + "epoch": 26.74425370963049, + "grad_norm": 0.06176143139600754, + "learning_rate": 4.27163649322289e-06, + "loss": 0.0114, + "step": 91920 + }, + { + "epoch": 26.747163223741634, + "grad_norm": 0.08122782409191132, + "learning_rate": 4.268747895943129e-06, + "loss": 0.0113, + "step": 91930 + }, + { + "epoch": 26.75007273785278, + "grad_norm": 0.10097743570804596, + "learning_rate": 4.265860113631142e-06, + "loss": 0.0105, + "step": 91940 + }, + { + "epoch": 26.75298225196392, + "grad_norm": 0.09584463387727737, + "learning_rate": 4.262973146506241e-06, + "loss": 0.0117, + "step": 91950 + }, + { + "epoch": 26.755891766075067, + "grad_norm": 0.11343914270401001, + "learning_rate": 4.260086994787669e-06, + "loss": 0.0111, + "step": 91960 + }, + { + "epoch": 26.75880128018621, + "grad_norm": 0.12820829451084137, + "learning_rate": 4.257201658694612e-06, + "loss": 0.0135, + "step": 91970 + }, + { + "epoch": 26.761710794297354, + "grad_norm": 0.1539735049009323, + "learning_rate": 4.2543171384461985e-06, + "loss": 0.0143, + "step": 91980 + }, + { + "epoch": 26.764620308408496, + "grad_norm": 0.11787393689155579, + "learning_rate": 4.251433434261477e-06, + "loss": 0.0106, + "step": 91990 + }, + { + "epoch": 26.767529822519638, + "grad_norm": 0.10539472848176956, + "learning_rate": 4.248550546359458e-06, + "loss": 0.011, + "step": 92000 + }, + { + "epoch": 26.770439336630783, + "grad_norm": 0.13080309331417084, + "learning_rate": 4.245668474959065e-06, + "loss": 0.0142, + "step": 92010 + }, + { + "epoch": 26.773348850741925, + "grad_norm": 0.26983627676963806, + "learning_rate": 4.242787220279185e-06, + "loss": 0.0125, + "step": 92020 + }, + { + "epoch": 26.77625836485307, + "grad_norm": 0.09890328347682953, + "learning_rate": 4.23990678253862e-06, + "loss": 0.0116, + "step": 92030 + }, + { + "epoch": 26.779167878964213, + "grad_norm": 0.11391404271125793, + "learning_rate": 4.237027161956122e-06, + "loss": 0.0117, + "step": 92040 + }, + { + "epoch": 26.782077393075355, + "grad_norm": 0.08783342689275742, + "learning_rate": 4.234148358750388e-06, + "loss": 0.0114, + "step": 92050 + }, + { + "epoch": 26.7849869071865, + "grad_norm": 0.1373830884695053, + "learning_rate": 4.231270373140037e-06, + "loss": 0.012, + "step": 92060 + }, + { + "epoch": 26.787896421297642, + "grad_norm": 0.11733800172805786, + "learning_rate": 4.228393205343626e-06, + "loss": 0.0135, + "step": 92070 + }, + { + "epoch": 26.790805935408788, + "grad_norm": 0.10977676510810852, + "learning_rate": 4.2255168555796715e-06, + "loss": 0.012, + "step": 92080 + }, + { + "epoch": 26.79371544951993, + "grad_norm": 0.08962979167699814, + "learning_rate": 4.2226413240666086e-06, + "loss": 0.011, + "step": 92090 + }, + { + "epoch": 26.796624963631075, + "grad_norm": 0.08938459306955338, + "learning_rate": 4.219766611022807e-06, + "loss": 0.0118, + "step": 92100 + }, + { + "epoch": 26.799534477742217, + "grad_norm": 0.0984835997223854, + "learning_rate": 4.216892716666588e-06, + "loss": 0.01, + "step": 92110 + }, + { + "epoch": 26.80244399185336, + "grad_norm": 0.10502730309963226, + "learning_rate": 4.214019641216209e-06, + "loss": 0.0118, + "step": 92120 + }, + { + "epoch": 26.805353505964504, + "grad_norm": 0.08048006147146225, + "learning_rate": 4.211147384889858e-06, + "loss": 0.0101, + "step": 92130 + }, + { + "epoch": 26.808263020075646, + "grad_norm": 0.11698340624570847, + "learning_rate": 4.208275947905659e-06, + "loss": 0.0123, + "step": 92140 + }, + { + "epoch": 26.81117253418679, + "grad_norm": 0.10749784857034683, + "learning_rate": 4.205405330481681e-06, + "loss": 0.0127, + "step": 92150 + }, + { + "epoch": 26.814082048297934, + "grad_norm": 0.12078143656253815, + "learning_rate": 4.202535532835934e-06, + "loss": 0.011, + "step": 92160 + }, + { + "epoch": 26.81699156240908, + "grad_norm": 0.11562138795852661, + "learning_rate": 4.199666555186354e-06, + "loss": 0.0127, + "step": 92170 + }, + { + "epoch": 26.81990107652022, + "grad_norm": 0.09818000346422195, + "learning_rate": 4.196798397750821e-06, + "loss": 0.0122, + "step": 92180 + }, + { + "epoch": 26.822810590631363, + "grad_norm": 0.12329849600791931, + "learning_rate": 4.1939310607471595e-06, + "loss": 0.0121, + "step": 92190 + }, + { + "epoch": 26.82572010474251, + "grad_norm": 0.08824770897626877, + "learning_rate": 4.191064544393122e-06, + "loss": 0.0144, + "step": 92200 + }, + { + "epoch": 26.82862961885365, + "grad_norm": 0.08534722030162811, + "learning_rate": 4.188198848906393e-06, + "loss": 0.0101, + "step": 92210 + }, + { + "epoch": 26.831539132964796, + "grad_norm": 0.14103595912456512, + "learning_rate": 4.185333974504611e-06, + "loss": 0.0112, + "step": 92220 + }, + { + "epoch": 26.834448647075938, + "grad_norm": 0.09276002645492554, + "learning_rate": 4.1824699214053474e-06, + "loss": 0.0102, + "step": 92230 + }, + { + "epoch": 26.837358161187083, + "grad_norm": 0.13936731219291687, + "learning_rate": 4.179606689826098e-06, + "loss": 0.0115, + "step": 92240 + }, + { + "epoch": 26.840267675298225, + "grad_norm": 0.12857003509998322, + "learning_rate": 4.176744279984319e-06, + "loss": 0.0123, + "step": 92250 + }, + { + "epoch": 26.843177189409367, + "grad_norm": 0.10333877801895142, + "learning_rate": 4.173882692097379e-06, + "loss": 0.0108, + "step": 92260 + }, + { + "epoch": 26.846086703520513, + "grad_norm": 0.08836212754249573, + "learning_rate": 4.171021926382605e-06, + "loss": 0.0095, + "step": 92270 + }, + { + "epoch": 26.848996217631655, + "grad_norm": 0.10284505039453506, + "learning_rate": 4.1681619830572484e-06, + "loss": 0.0116, + "step": 92280 + }, + { + "epoch": 26.8519057317428, + "grad_norm": 0.11580946296453476, + "learning_rate": 4.165302862338504e-06, + "loss": 0.0098, + "step": 92290 + }, + { + "epoch": 26.854815245853942, + "grad_norm": 0.1068500280380249, + "learning_rate": 4.1624445644435094e-06, + "loss": 0.0145, + "step": 92300 + }, + { + "epoch": 26.857724759965087, + "grad_norm": 0.473379909992218, + "learning_rate": 4.159587089589325e-06, + "loss": 0.0108, + "step": 92310 + }, + { + "epoch": 26.86063427407623, + "grad_norm": 0.11877524852752686, + "learning_rate": 4.156730437992965e-06, + "loss": 0.0114, + "step": 92320 + }, + { + "epoch": 26.86354378818737, + "grad_norm": 0.10467761754989624, + "learning_rate": 4.153874609871364e-06, + "loss": 0.0109, + "step": 92330 + }, + { + "epoch": 26.866453302298517, + "grad_norm": 0.13698382675647736, + "learning_rate": 4.151019605441406e-06, + "loss": 0.0119, + "step": 92340 + }, + { + "epoch": 26.86936281640966, + "grad_norm": 0.095960333943367, + "learning_rate": 4.148165424919919e-06, + "loss": 0.0122, + "step": 92350 + }, + { + "epoch": 26.872272330520804, + "grad_norm": 0.22270652651786804, + "learning_rate": 4.145312068523652e-06, + "loss": 0.011, + "step": 92360 + }, + { + "epoch": 26.875181844631946, + "grad_norm": 0.07740391790866852, + "learning_rate": 4.142459536469292e-06, + "loss": 0.0108, + "step": 92370 + }, + { + "epoch": 26.87809135874309, + "grad_norm": 0.07714121043682098, + "learning_rate": 4.139607828973479e-06, + "loss": 0.0123, + "step": 92380 + }, + { + "epoch": 26.881000872854234, + "grad_norm": 0.1104799434542656, + "learning_rate": 4.136756946252784e-06, + "loss": 0.0116, + "step": 92390 + }, + { + "epoch": 26.883910386965375, + "grad_norm": 0.1066170334815979, + "learning_rate": 4.133906888523702e-06, + "loss": 0.0121, + "step": 92400 + }, + { + "epoch": 26.88681990107652, + "grad_norm": 0.09247691184282303, + "learning_rate": 4.1310576560026816e-06, + "loss": 0.0117, + "step": 92410 + }, + { + "epoch": 26.889729415187663, + "grad_norm": 0.09868945181369781, + "learning_rate": 4.128209248906108e-06, + "loss": 0.0123, + "step": 92420 + }, + { + "epoch": 26.89263892929881, + "grad_norm": 0.10409216582775116, + "learning_rate": 4.125361667450295e-06, + "loss": 0.013, + "step": 92430 + }, + { + "epoch": 26.89554844340995, + "grad_norm": 0.09770956635475159, + "learning_rate": 4.122514911851491e-06, + "loss": 0.0138, + "step": 92440 + }, + { + "epoch": 26.898457957521096, + "grad_norm": 0.10641931742429733, + "learning_rate": 4.119668982325897e-06, + "loss": 0.0118, + "step": 92450 + }, + { + "epoch": 26.901367471632238, + "grad_norm": 0.11176233738660812, + "learning_rate": 4.116823879089642e-06, + "loss": 0.0097, + "step": 92460 + }, + { + "epoch": 26.90427698574338, + "grad_norm": 0.1121622622013092, + "learning_rate": 4.113979602358792e-06, + "loss": 0.0123, + "step": 92470 + }, + { + "epoch": 26.907186499854525, + "grad_norm": 0.10391731560230255, + "learning_rate": 4.11113615234934e-06, + "loss": 0.011, + "step": 92480 + }, + { + "epoch": 26.910096013965667, + "grad_norm": 0.08893290162086487, + "learning_rate": 4.108293529277245e-06, + "loss": 0.0109, + "step": 92490 + }, + { + "epoch": 26.913005528076813, + "grad_norm": 0.1407746821641922, + "learning_rate": 4.105451733358379e-06, + "loss": 0.0113, + "step": 92500 + }, + { + "epoch": 26.915915042187954, + "grad_norm": 0.09385742247104645, + "learning_rate": 4.102610764808551e-06, + "loss": 0.0108, + "step": 92510 + }, + { + "epoch": 26.918824556299096, + "grad_norm": 0.1335674226284027, + "learning_rate": 4.099770623843519e-06, + "loss": 0.0107, + "step": 92520 + }, + { + "epoch": 26.921734070410242, + "grad_norm": 0.07395434379577637, + "learning_rate": 4.096931310678977e-06, + "loss": 0.0112, + "step": 92530 + }, + { + "epoch": 26.924643584521384, + "grad_norm": 0.11884579062461853, + "learning_rate": 4.094092825530544e-06, + "loss": 0.0135, + "step": 92540 + }, + { + "epoch": 26.92755309863253, + "grad_norm": 0.08266336470842361, + "learning_rate": 4.091255168613792e-06, + "loss": 0.0122, + "step": 92550 + }, + { + "epoch": 26.93046261274367, + "grad_norm": 0.10173249989748001, + "learning_rate": 4.0884183401442155e-06, + "loss": 0.0117, + "step": 92560 + }, + { + "epoch": 26.933372126854817, + "grad_norm": 0.1460864245891571, + "learning_rate": 4.08558234033726e-06, + "loss": 0.012, + "step": 92570 + }, + { + "epoch": 26.93628164096596, + "grad_norm": 0.07080257683992386, + "learning_rate": 4.082747169408292e-06, + "loss": 0.0101, + "step": 92580 + }, + { + "epoch": 26.9391911550771, + "grad_norm": 0.10333485156297684, + "learning_rate": 4.079912827572629e-06, + "loss": 0.0131, + "step": 92590 + }, + { + "epoch": 26.942100669188246, + "grad_norm": 0.0816969946026802, + "learning_rate": 4.077079315045524e-06, + "loss": 0.0091, + "step": 92600 + }, + { + "epoch": 26.945010183299388, + "grad_norm": 0.15164759755134583, + "learning_rate": 4.0742466320421565e-06, + "loss": 0.012, + "step": 92610 + }, + { + "epoch": 26.947919697410534, + "grad_norm": 0.1259452998638153, + "learning_rate": 4.071414778777657e-06, + "loss": 0.0098, + "step": 92620 + }, + { + "epoch": 26.950829211521675, + "grad_norm": 0.11392171680927277, + "learning_rate": 4.06858375546708e-06, + "loss": 0.0108, + "step": 92630 + }, + { + "epoch": 26.95373872563282, + "grad_norm": 0.06921622902154922, + "learning_rate": 4.065753562325428e-06, + "loss": 0.0107, + "step": 92640 + }, + { + "epoch": 26.956648239743963, + "grad_norm": 0.19174450635910034, + "learning_rate": 4.062924199567631e-06, + "loss": 0.0121, + "step": 92650 + }, + { + "epoch": 26.959557753855105, + "grad_norm": 0.12076525390148163, + "learning_rate": 4.060095667408566e-06, + "loss": 0.0114, + "step": 92660 + }, + { + "epoch": 26.96246726796625, + "grad_norm": 0.10767622292041779, + "learning_rate": 4.057267966063034e-06, + "loss": 0.0125, + "step": 92670 + }, + { + "epoch": 26.965376782077392, + "grad_norm": 0.15069574117660522, + "learning_rate": 4.054441095745782e-06, + "loss": 0.012, + "step": 92680 + }, + { + "epoch": 26.968286296188538, + "grad_norm": 0.14754079282283783, + "learning_rate": 4.0516150566715005e-06, + "loss": 0.0115, + "step": 92690 + }, + { + "epoch": 26.97119581029968, + "grad_norm": 0.09692368656396866, + "learning_rate": 4.048789849054798e-06, + "loss": 0.0118, + "step": 92700 + }, + { + "epoch": 26.974105324410825, + "grad_norm": 0.10250865668058395, + "learning_rate": 4.045965473110235e-06, + "loss": 0.0125, + "step": 92710 + }, + { + "epoch": 26.977014838521967, + "grad_norm": 0.1325760781764984, + "learning_rate": 4.043141929052309e-06, + "loss": 0.0132, + "step": 92720 + }, + { + "epoch": 26.97992435263311, + "grad_norm": 0.0843382179737091, + "learning_rate": 4.0403192170954435e-06, + "loss": 0.0134, + "step": 92730 + }, + { + "epoch": 26.982833866744254, + "grad_norm": 0.11270768195390701, + "learning_rate": 4.037497337454001e-06, + "loss": 0.0115, + "step": 92740 + }, + { + "epoch": 26.985743380855396, + "grad_norm": 0.18487466871738434, + "learning_rate": 4.034676290342289e-06, + "loss": 0.0108, + "step": 92750 + }, + { + "epoch": 26.988652894966542, + "grad_norm": 0.10621064901351929, + "learning_rate": 4.031856075974555e-06, + "loss": 0.0118, + "step": 92760 + }, + { + "epoch": 26.991562409077684, + "grad_norm": 0.10340508818626404, + "learning_rate": 4.029036694564968e-06, + "loss": 0.0115, + "step": 92770 + }, + { + "epoch": 26.994471923188826, + "grad_norm": 0.08859309554100037, + "learning_rate": 4.0262181463276316e-06, + "loss": 0.0097, + "step": 92780 + }, + { + "epoch": 26.99738143729997, + "grad_norm": 0.1010841578245163, + "learning_rate": 4.023400431476615e-06, + "loss": 0.0123, + "step": 92790 + }, + { + "epoch": 27.000290951411113, + "grad_norm": 0.0801166519522667, + "learning_rate": 4.020583550225896e-06, + "loss": 0.013, + "step": 92800 + }, + { + "epoch": 27.00320046552226, + "grad_norm": 0.17988455295562744, + "learning_rate": 4.0177675027893965e-06, + "loss": 0.0114, + "step": 92810 + }, + { + "epoch": 27.0061099796334, + "grad_norm": 0.12971754372119904, + "learning_rate": 4.014952289380977e-06, + "loss": 0.0122, + "step": 92820 + }, + { + "epoch": 27.009019493744546, + "grad_norm": 0.09873839467763901, + "learning_rate": 4.012137910214441e-06, + "loss": 0.0106, + "step": 92830 + }, + { + "epoch": 27.011929007855688, + "grad_norm": 0.08898330479860306, + "learning_rate": 4.009324365503516e-06, + "loss": 0.0107, + "step": 92840 + }, + { + "epoch": 27.01483852196683, + "grad_norm": 0.1646529585123062, + "learning_rate": 4.006511655461871e-06, + "loss": 0.0122, + "step": 92850 + }, + { + "epoch": 27.017748036077975, + "grad_norm": 0.09366992861032486, + "learning_rate": 4.003699780303114e-06, + "loss": 0.0133, + "step": 92860 + }, + { + "epoch": 27.020657550189117, + "grad_norm": 0.08683478832244873, + "learning_rate": 4.000888740240793e-06, + "loss": 0.011, + "step": 92870 + }, + { + "epoch": 27.023567064300263, + "grad_norm": 0.29735320806503296, + "learning_rate": 3.998078535488382e-06, + "loss": 0.0117, + "step": 92880 + }, + { + "epoch": 27.026476578411405, + "grad_norm": 0.1347411721944809, + "learning_rate": 3.9952691662592995e-06, + "loss": 0.0114, + "step": 92890 + }, + { + "epoch": 27.02938609252255, + "grad_norm": 0.08631578832864761, + "learning_rate": 3.992460632766902e-06, + "loss": 0.0111, + "step": 92900 + }, + { + "epoch": 27.032295606633692, + "grad_norm": 0.1330031305551529, + "learning_rate": 3.9896529352244715e-06, + "loss": 0.0137, + "step": 92910 + }, + { + "epoch": 27.035205120744834, + "grad_norm": 0.08556458353996277, + "learning_rate": 3.986846073845243e-06, + "loss": 0.0129, + "step": 92920 + }, + { + "epoch": 27.03811463485598, + "grad_norm": 0.17518799006938934, + "learning_rate": 3.984040048842372e-06, + "loss": 0.0127, + "step": 92930 + }, + { + "epoch": 27.04102414896712, + "grad_norm": 0.12118197232484818, + "learning_rate": 3.981234860428963e-06, + "loss": 0.0104, + "step": 92940 + }, + { + "epoch": 27.043933663078267, + "grad_norm": 0.11654990911483765, + "learning_rate": 3.978430508818044e-06, + "loss": 0.0122, + "step": 92950 + }, + { + "epoch": 27.04684317718941, + "grad_norm": 0.09441301971673965, + "learning_rate": 3.9756269942225985e-06, + "loss": 0.0124, + "step": 92960 + }, + { + "epoch": 27.049752691300554, + "grad_norm": 0.12860102951526642, + "learning_rate": 3.972824316855524e-06, + "loss": 0.0099, + "step": 92970 + }, + { + "epoch": 27.052662205411696, + "grad_norm": 0.09998370707035065, + "learning_rate": 3.970022476929667e-06, + "loss": 0.0113, + "step": 92980 + }, + { + "epoch": 27.05557171952284, + "grad_norm": 0.13079224526882172, + "learning_rate": 3.9672214746578165e-06, + "loss": 0.0133, + "step": 92990 + }, + { + "epoch": 27.058481233633984, + "grad_norm": 0.07109879702329636, + "learning_rate": 3.964421310252682e-06, + "loss": 0.0125, + "step": 93000 + }, + { + "epoch": 27.061390747745126, + "grad_norm": 0.09731003642082214, + "learning_rate": 3.961621983926923e-06, + "loss": 0.0116, + "step": 93010 + }, + { + "epoch": 27.06430026185627, + "grad_norm": 0.09186958521604538, + "learning_rate": 3.958823495893124e-06, + "loss": 0.0116, + "step": 93020 + }, + { + "epoch": 27.067209775967413, + "grad_norm": 0.15172959864139557, + "learning_rate": 3.95602584636382e-06, + "loss": 0.0108, + "step": 93030 + }, + { + "epoch": 27.07011929007856, + "grad_norm": 0.09289747476577759, + "learning_rate": 3.953229035551464e-06, + "loss": 0.0117, + "step": 93040 + }, + { + "epoch": 27.0730288041897, + "grad_norm": 0.07643039524555206, + "learning_rate": 3.95043306366846e-06, + "loss": 0.0109, + "step": 93050 + }, + { + "epoch": 27.075938318300842, + "grad_norm": 0.10168536752462387, + "learning_rate": 3.94763793092715e-06, + "loss": 0.0125, + "step": 93060 + }, + { + "epoch": 27.078847832411988, + "grad_norm": 0.09128133952617645, + "learning_rate": 3.944843637539797e-06, + "loss": 0.0116, + "step": 93070 + }, + { + "epoch": 27.08175734652313, + "grad_norm": 0.08423581719398499, + "learning_rate": 3.942050183718606e-06, + "loss": 0.0113, + "step": 93080 + }, + { + "epoch": 27.084666860634275, + "grad_norm": 0.09448714554309845, + "learning_rate": 3.939257569675735e-06, + "loss": 0.0107, + "step": 93090 + }, + { + "epoch": 27.087576374745417, + "grad_norm": 0.13805058598518372, + "learning_rate": 3.936465795623257e-06, + "loss": 0.0112, + "step": 93100 + }, + { + "epoch": 27.090485888856563, + "grad_norm": 0.11928583681583405, + "learning_rate": 3.933674861773185e-06, + "loss": 0.0118, + "step": 93110 + }, + { + "epoch": 27.093395402967705, + "grad_norm": 0.12881624698638916, + "learning_rate": 3.930884768337476e-06, + "loss": 0.0134, + "step": 93120 + }, + { + "epoch": 27.096304917078847, + "grad_norm": 0.09364187717437744, + "learning_rate": 3.928095515528022e-06, + "loss": 0.0107, + "step": 93130 + }, + { + "epoch": 27.099214431189992, + "grad_norm": 0.34514328837394714, + "learning_rate": 3.9253071035566476e-06, + "loss": 0.0106, + "step": 93140 + }, + { + "epoch": 27.102123945301134, + "grad_norm": 0.08505402505397797, + "learning_rate": 3.9225195326351056e-06, + "loss": 0.0109, + "step": 93150 + }, + { + "epoch": 27.10503345941228, + "grad_norm": 0.13663069903850555, + "learning_rate": 3.919732802975102e-06, + "loss": 0.012, + "step": 93160 + }, + { + "epoch": 27.10794297352342, + "grad_norm": 0.12823836505413055, + "learning_rate": 3.916946914788272e-06, + "loss": 0.012, + "step": 93170 + }, + { + "epoch": 27.110852487634563, + "grad_norm": 0.127447247505188, + "learning_rate": 3.914161868286177e-06, + "loss": 0.01, + "step": 93180 + }, + { + "epoch": 27.11376200174571, + "grad_norm": 0.12690603733062744, + "learning_rate": 3.9113776636803296e-06, + "loss": 0.0112, + "step": 93190 + }, + { + "epoch": 27.11667151585685, + "grad_norm": 0.09973982721567154, + "learning_rate": 3.908594301182172e-06, + "loss": 0.0109, + "step": 93200 + }, + { + "epoch": 27.119581029967996, + "grad_norm": 0.1418268382549286, + "learning_rate": 3.905811781003083e-06, + "loss": 0.0107, + "step": 93210 + }, + { + "epoch": 27.122490544079138, + "grad_norm": 0.12688018381595612, + "learning_rate": 3.9030301033543674e-06, + "loss": 0.0102, + "step": 93220 + }, + { + "epoch": 27.125400058190284, + "grad_norm": 0.1264631599187851, + "learning_rate": 3.900249268447283e-06, + "loss": 0.0106, + "step": 93230 + }, + { + "epoch": 27.128309572301426, + "grad_norm": 0.10690052062273026, + "learning_rate": 3.897469276493018e-06, + "loss": 0.0109, + "step": 93240 + }, + { + "epoch": 27.131219086412568, + "grad_norm": 0.12037681043148041, + "learning_rate": 3.894690127702689e-06, + "loss": 0.0113, + "step": 93250 + }, + { + "epoch": 27.134128600523713, + "grad_norm": 0.11801845580339432, + "learning_rate": 3.891911822287359e-06, + "loss": 0.0102, + "step": 93260 + }, + { + "epoch": 27.137038114634855, + "grad_norm": 0.10939665138721466, + "learning_rate": 3.889134360458014e-06, + "loss": 0.0096, + "step": 93270 + }, + { + "epoch": 27.139947628746, + "grad_norm": 0.08864489197731018, + "learning_rate": 3.88635774242559e-06, + "loss": 0.0139, + "step": 93280 + }, + { + "epoch": 27.142857142857142, + "grad_norm": 0.07263367623090744, + "learning_rate": 3.883581968400957e-06, + "loss": 0.0111, + "step": 93290 + }, + { + "epoch": 27.145766656968288, + "grad_norm": 0.15792590379714966, + "learning_rate": 3.8808070385949065e-06, + "loss": 0.0102, + "step": 93300 + }, + { + "epoch": 27.14867617107943, + "grad_norm": 0.09783598780632019, + "learning_rate": 3.878032953218183e-06, + "loss": 0.0112, + "step": 93310 + }, + { + "epoch": 27.15158568519057, + "grad_norm": 0.0786840096116066, + "learning_rate": 3.875259712481457e-06, + "loss": 0.0143, + "step": 93320 + }, + { + "epoch": 27.154495199301717, + "grad_norm": 0.13996651768684387, + "learning_rate": 3.872487316595342e-06, + "loss": 0.0123, + "step": 93330 + }, + { + "epoch": 27.15740471341286, + "grad_norm": 0.15401732921600342, + "learning_rate": 3.8697157657703755e-06, + "loss": 0.0137, + "step": 93340 + }, + { + "epoch": 27.160314227524005, + "grad_norm": 0.10242515057325363, + "learning_rate": 3.866945060217044e-06, + "loss": 0.0117, + "step": 93350 + }, + { + "epoch": 27.163223741635147, + "grad_norm": 0.11108355969190598, + "learning_rate": 3.864175200145767e-06, + "loss": 0.0144, + "step": 93360 + }, + { + "epoch": 27.166133255746292, + "grad_norm": 0.08687623590230942, + "learning_rate": 3.861406185766894e-06, + "loss": 0.0095, + "step": 93370 + }, + { + "epoch": 27.169042769857434, + "grad_norm": 0.07990217208862305, + "learning_rate": 3.858638017290704e-06, + "loss": 0.0102, + "step": 93380 + }, + { + "epoch": 27.171952283968576, + "grad_norm": 0.10726264119148254, + "learning_rate": 3.855870694927439e-06, + "loss": 0.0094, + "step": 93390 + }, + { + "epoch": 27.17486179807972, + "grad_norm": 0.09072327613830566, + "learning_rate": 3.853104218887249e-06, + "loss": 0.0114, + "step": 93400 + }, + { + "epoch": 27.177771312190863, + "grad_norm": 0.09515589475631714, + "learning_rate": 3.8503385893802275e-06, + "loss": 0.0111, + "step": 93410 + }, + { + "epoch": 27.18068082630201, + "grad_norm": 0.10817790031433105, + "learning_rate": 3.847573806616408e-06, + "loss": 0.0115, + "step": 93420 + }, + { + "epoch": 27.18359034041315, + "grad_norm": 0.13244600594043732, + "learning_rate": 3.8448098708057626e-06, + "loss": 0.0156, + "step": 93430 + }, + { + "epoch": 27.186499854524296, + "grad_norm": 0.12518276274204254, + "learning_rate": 3.842046782158189e-06, + "loss": 0.0108, + "step": 93440 + }, + { + "epoch": 27.189409368635438, + "grad_norm": 0.12219314277172089, + "learning_rate": 3.839284540883522e-06, + "loss": 0.0127, + "step": 93450 + }, + { + "epoch": 27.19231888274658, + "grad_norm": 0.15441487729549408, + "learning_rate": 3.836523147191541e-06, + "loss": 0.0113, + "step": 93460 + }, + { + "epoch": 27.195228396857726, + "grad_norm": 0.06515106558799744, + "learning_rate": 3.833762601291958e-06, + "loss": 0.0116, + "step": 93470 + }, + { + "epoch": 27.198137910968867, + "grad_norm": 0.0819002017378807, + "learning_rate": 3.831002903394414e-06, + "loss": 0.0119, + "step": 93480 + }, + { + "epoch": 27.201047425080013, + "grad_norm": 0.11505831032991409, + "learning_rate": 3.828244053708484e-06, + "loss": 0.012, + "step": 93490 + }, + { + "epoch": 27.203956939191155, + "grad_norm": 0.11051633954048157, + "learning_rate": 3.825486052443697e-06, + "loss": 0.0095, + "step": 93500 + }, + { + "epoch": 27.206866453302297, + "grad_norm": 0.09388643503189087, + "learning_rate": 3.822728899809501e-06, + "loss": 0.0132, + "step": 93510 + }, + { + "epoch": 27.209775967413442, + "grad_norm": 0.10651973634958267, + "learning_rate": 3.819972596015276e-06, + "loss": 0.0093, + "step": 93520 + }, + { + "epoch": 27.212685481524584, + "grad_norm": 0.08555954694747925, + "learning_rate": 3.817217141270351e-06, + "loss": 0.01, + "step": 93530 + }, + { + "epoch": 27.21559499563573, + "grad_norm": 0.11285668611526489, + "learning_rate": 3.8144625357839885e-06, + "loss": 0.0112, + "step": 93540 + }, + { + "epoch": 27.21850450974687, + "grad_norm": 0.09565701335668564, + "learning_rate": 3.8117087797653727e-06, + "loss": 0.0123, + "step": 93550 + }, + { + "epoch": 27.221414023858017, + "grad_norm": 0.10066480934619904, + "learning_rate": 3.808955873423646e-06, + "loss": 0.0128, + "step": 93560 + }, + { + "epoch": 27.22432353796916, + "grad_norm": 0.17633379995822906, + "learning_rate": 3.8062038169678588e-06, + "loss": 0.0127, + "step": 93570 + }, + { + "epoch": 27.2272330520803, + "grad_norm": 0.09696795791387558, + "learning_rate": 3.8034526106070244e-06, + "loss": 0.0111, + "step": 93580 + }, + { + "epoch": 27.230142566191446, + "grad_norm": 0.13921082019805908, + "learning_rate": 3.8007022545500706e-06, + "loss": 0.0109, + "step": 93590 + }, + { + "epoch": 27.23305208030259, + "grad_norm": 0.10837170481681824, + "learning_rate": 3.7979527490058707e-06, + "loss": 0.0111, + "step": 93600 + }, + { + "epoch": 27.235961594413734, + "grad_norm": 0.14415352046489716, + "learning_rate": 3.7952040941832356e-06, + "loss": 0.0087, + "step": 93610 + }, + { + "epoch": 27.238871108524876, + "grad_norm": 0.13389325141906738, + "learning_rate": 3.7924562902909018e-06, + "loss": 0.0125, + "step": 93620 + }, + { + "epoch": 27.24178062263602, + "grad_norm": 0.09202083200216293, + "learning_rate": 3.789709337537552e-06, + "loss": 0.0116, + "step": 93630 + }, + { + "epoch": 27.244690136747163, + "grad_norm": 0.07119566202163696, + "learning_rate": 3.7869632361317928e-06, + "loss": 0.012, + "step": 93640 + }, + { + "epoch": 27.247599650858305, + "grad_norm": 0.07909061014652252, + "learning_rate": 3.784217986282178e-06, + "loss": 0.0104, + "step": 93650 + }, + { + "epoch": 27.25050916496945, + "grad_norm": 0.11125337332487106, + "learning_rate": 3.781473588197192e-06, + "loss": 0.01, + "step": 93660 + }, + { + "epoch": 27.253418679080593, + "grad_norm": 0.11387079209089279, + "learning_rate": 3.778730042085251e-06, + "loss": 0.0115, + "step": 93670 + }, + { + "epoch": 27.256328193191738, + "grad_norm": 0.08182413130998611, + "learning_rate": 3.7759873481547057e-06, + "loss": 0.0123, + "step": 93680 + }, + { + "epoch": 27.25923770730288, + "grad_norm": 0.08462554961442947, + "learning_rate": 3.7732455066138493e-06, + "loss": 0.0132, + "step": 93690 + }, + { + "epoch": 27.262147221414025, + "grad_norm": 0.1540762186050415, + "learning_rate": 3.77050451767091e-06, + "loss": 0.0121, + "step": 93700 + }, + { + "epoch": 27.265056735525167, + "grad_norm": 0.1305098682641983, + "learning_rate": 3.7677643815340403e-06, + "loss": 0.0106, + "step": 93710 + }, + { + "epoch": 27.26796624963631, + "grad_norm": 0.08059406280517578, + "learning_rate": 3.7650250984113393e-06, + "loss": 0.0114, + "step": 93720 + }, + { + "epoch": 27.270875763747455, + "grad_norm": 0.12834946811199188, + "learning_rate": 3.7622866685108433e-06, + "loss": 0.0111, + "step": 93730 + }, + { + "epoch": 27.273785277858597, + "grad_norm": 0.14321938157081604, + "learning_rate": 3.759549092040512e-06, + "loss": 0.012, + "step": 93740 + }, + { + "epoch": 27.276694791969742, + "grad_norm": 0.11801843345165253, + "learning_rate": 3.7568123692082423e-06, + "loss": 0.0114, + "step": 93750 + }, + { + "epoch": 27.279604306080884, + "grad_norm": 0.1255379319190979, + "learning_rate": 3.7540765002218736e-06, + "loss": 0.0124, + "step": 93760 + }, + { + "epoch": 27.28251382019203, + "grad_norm": 0.1803712099790573, + "learning_rate": 3.7513414852891826e-06, + "loss": 0.011, + "step": 93770 + }, + { + "epoch": 27.28542333430317, + "grad_norm": 0.0934293270111084, + "learning_rate": 3.7486073246178714e-06, + "loss": 0.0109, + "step": 93780 + }, + { + "epoch": 27.288332848414314, + "grad_norm": 0.09717763215303421, + "learning_rate": 3.7458740184155726e-06, + "loss": 0.011, + "step": 93790 + }, + { + "epoch": 27.29124236252546, + "grad_norm": 0.10298573225736618, + "learning_rate": 3.74314156688988e-06, + "loss": 0.011, + "step": 93800 + }, + { + "epoch": 27.2941518766366, + "grad_norm": 0.06576182693243027, + "learning_rate": 3.7404099702482973e-06, + "loss": 0.009, + "step": 93810 + }, + { + "epoch": 27.297061390747746, + "grad_norm": 0.10978743433952332, + "learning_rate": 3.737679228698266e-06, + "loss": 0.0112, + "step": 93820 + }, + { + "epoch": 27.29997090485889, + "grad_norm": 0.0954269990324974, + "learning_rate": 3.7349493424471708e-06, + "loss": 0.013, + "step": 93830 + }, + { + "epoch": 27.30288041897003, + "grad_norm": 0.13360445201396942, + "learning_rate": 3.732220311702337e-06, + "loss": 0.0113, + "step": 93840 + }, + { + "epoch": 27.305789933081176, + "grad_norm": 0.07584881037473679, + "learning_rate": 3.729492136671008e-06, + "loss": 0.0116, + "step": 93850 + }, + { + "epoch": 27.308699447192318, + "grad_norm": 0.1658412218093872, + "learning_rate": 3.726764817560368e-06, + "loss": 0.0108, + "step": 93860 + }, + { + "epoch": 27.311608961303463, + "grad_norm": 0.0883476734161377, + "learning_rate": 3.7240383545775437e-06, + "loss": 0.0098, + "step": 93870 + }, + { + "epoch": 27.314518475414605, + "grad_norm": 0.09576579183340073, + "learning_rate": 3.7213127479295954e-06, + "loss": 0.0112, + "step": 93880 + }, + { + "epoch": 27.31742798952575, + "grad_norm": 0.09901919215917587, + "learning_rate": 3.7185879978235064e-06, + "loss": 0.0118, + "step": 93890 + }, + { + "epoch": 27.320337503636893, + "grad_norm": 0.07786305993795395, + "learning_rate": 3.7158641044662096e-06, + "loss": 0.0114, + "step": 93900 + }, + { + "epoch": 27.323247017748034, + "grad_norm": 0.09601613879203796, + "learning_rate": 3.713141068064569e-06, + "loss": 0.013, + "step": 93910 + }, + { + "epoch": 27.32615653185918, + "grad_norm": 0.10581165552139282, + "learning_rate": 3.7104188888253726e-06, + "loss": 0.0107, + "step": 93920 + }, + { + "epoch": 27.329066045970322, + "grad_norm": 0.14208471775054932, + "learning_rate": 3.7076975669553604e-06, + "loss": 0.0121, + "step": 93930 + }, + { + "epoch": 27.331975560081467, + "grad_norm": 0.1389925628900528, + "learning_rate": 3.7049771026611916e-06, + "loss": 0.0127, + "step": 93940 + }, + { + "epoch": 27.33488507419261, + "grad_norm": 0.1001415029168129, + "learning_rate": 3.7022574961494764e-06, + "loss": 0.0106, + "step": 93950 + }, + { + "epoch": 27.337794588303755, + "grad_norm": 0.12371004372835159, + "learning_rate": 3.6995387476267412e-06, + "loss": 0.0101, + "step": 93960 + }, + { + "epoch": 27.340704102414897, + "grad_norm": 0.1526404768228531, + "learning_rate": 3.6968208572994656e-06, + "loss": 0.0112, + "step": 93970 + }, + { + "epoch": 27.34361361652604, + "grad_norm": 0.05377037078142166, + "learning_rate": 3.694103825374046e-06, + "loss": 0.0109, + "step": 93980 + }, + { + "epoch": 27.346523130637184, + "grad_norm": 0.08587592095136642, + "learning_rate": 3.6913876520568296e-06, + "loss": 0.0108, + "step": 93990 + }, + { + "epoch": 27.349432644748326, + "grad_norm": 0.11606904864311218, + "learning_rate": 3.6886723375540958e-06, + "loss": 0.0115, + "step": 94000 + }, + { + "epoch": 27.35234215885947, + "grad_norm": 0.09281585365533829, + "learning_rate": 3.6859578820720434e-06, + "loss": 0.0105, + "step": 94010 + }, + { + "epoch": 27.355251672970613, + "grad_norm": 0.10341320186853409, + "learning_rate": 3.683244285816826e-06, + "loss": 0.011, + "step": 94020 + }, + { + "epoch": 27.35816118708176, + "grad_norm": 0.08330871164798737, + "learning_rate": 3.6805315489945245e-06, + "loss": 0.0096, + "step": 94030 + }, + { + "epoch": 27.3610707011929, + "grad_norm": 0.09992986917495728, + "learning_rate": 3.6778196718111485e-06, + "loss": 0.0106, + "step": 94040 + }, + { + "epoch": 27.363980215304043, + "grad_norm": 0.10565102845430374, + "learning_rate": 3.6751086544726457e-06, + "loss": 0.0155, + "step": 94050 + }, + { + "epoch": 27.36688972941519, + "grad_norm": 0.14161178469657898, + "learning_rate": 3.6723984971849024e-06, + "loss": 0.0126, + "step": 94060 + }, + { + "epoch": 27.36979924352633, + "grad_norm": 0.09731929749250412, + "learning_rate": 3.669689200153741e-06, + "loss": 0.0126, + "step": 94070 + }, + { + "epoch": 27.372708757637476, + "grad_norm": 0.09061943739652634, + "learning_rate": 3.666980763584912e-06, + "loss": 0.0123, + "step": 94080 + }, + { + "epoch": 27.375618271748618, + "grad_norm": 0.10220864415168762, + "learning_rate": 3.664273187684094e-06, + "loss": 0.0107, + "step": 94090 + }, + { + "epoch": 27.378527785859763, + "grad_norm": 0.07552880793809891, + "learning_rate": 3.661566472656927e-06, + "loss": 0.0134, + "step": 94100 + }, + { + "epoch": 27.381437299970905, + "grad_norm": 0.09306006133556366, + "learning_rate": 3.6588606187089597e-06, + "loss": 0.0108, + "step": 94110 + }, + { + "epoch": 27.384346814082047, + "grad_norm": 0.10749334841966629, + "learning_rate": 3.6561556260456775e-06, + "loss": 0.0098, + "step": 94120 + }, + { + "epoch": 27.387256328193192, + "grad_norm": 0.11530201882123947, + "learning_rate": 3.6534514948725145e-06, + "loss": 0.0117, + "step": 94130 + }, + { + "epoch": 27.390165842304334, + "grad_norm": 0.15295293927192688, + "learning_rate": 3.6507482253948334e-06, + "loss": 0.0107, + "step": 94140 + }, + { + "epoch": 27.39307535641548, + "grad_norm": 0.10071603953838348, + "learning_rate": 3.6480458178179244e-06, + "loss": 0.0115, + "step": 94150 + }, + { + "epoch": 27.395984870526622, + "grad_norm": 0.13424327969551086, + "learning_rate": 3.6453442723470173e-06, + "loss": 0.0102, + "step": 94160 + }, + { + "epoch": 27.398894384637764, + "grad_norm": 0.09991291165351868, + "learning_rate": 3.6426435891872774e-06, + "loss": 0.0103, + "step": 94170 + }, + { + "epoch": 27.40180389874891, + "grad_norm": 0.08748897165060043, + "learning_rate": 3.639943768543809e-06, + "loss": 0.0121, + "step": 94180 + }, + { + "epoch": 27.40471341286005, + "grad_norm": 0.11095276474952698, + "learning_rate": 3.6372448106216373e-06, + "loss": 0.0107, + "step": 94190 + }, + { + "epoch": 27.407622926971197, + "grad_norm": 0.1247183158993721, + "learning_rate": 3.634546715625734e-06, + "loss": 0.0119, + "step": 94200 + }, + { + "epoch": 27.41053244108234, + "grad_norm": 0.11863432824611664, + "learning_rate": 3.631849483761007e-06, + "loss": 0.0122, + "step": 94210 + }, + { + "epoch": 27.413441955193484, + "grad_norm": 0.07958199828863144, + "learning_rate": 3.629153115232285e-06, + "loss": 0.0092, + "step": 94220 + }, + { + "epoch": 27.416351469304626, + "grad_norm": 0.08630666136741638, + "learning_rate": 3.6264576102443462e-06, + "loss": 0.011, + "step": 94230 + }, + { + "epoch": 27.419260983415768, + "grad_norm": 0.11470384895801544, + "learning_rate": 3.62376296900189e-06, + "loss": 0.0107, + "step": 94240 + }, + { + "epoch": 27.422170497526913, + "grad_norm": 0.06836704164743423, + "learning_rate": 3.6210691917095613e-06, + "loss": 0.0125, + "step": 94250 + }, + { + "epoch": 27.425080011638055, + "grad_norm": 0.1486886739730835, + "learning_rate": 3.6183762785719303e-06, + "loss": 0.0115, + "step": 94260 + }, + { + "epoch": 27.4279895257492, + "grad_norm": 0.10942312330007553, + "learning_rate": 3.615684229793514e-06, + "loss": 0.012, + "step": 94270 + }, + { + "epoch": 27.430899039860343, + "grad_norm": 0.1345135122537613, + "learning_rate": 3.6129930455787446e-06, + "loss": 0.0123, + "step": 94280 + }, + { + "epoch": 27.433808553971488, + "grad_norm": 0.11267916858196259, + "learning_rate": 3.6103027261320075e-06, + "loss": 0.012, + "step": 94290 + }, + { + "epoch": 27.43671806808263, + "grad_norm": 0.12139949947595596, + "learning_rate": 3.607613271657616e-06, + "loss": 0.011, + "step": 94300 + }, + { + "epoch": 27.439627582193772, + "grad_norm": 0.1378583014011383, + "learning_rate": 3.60492468235981e-06, + "loss": 0.0116, + "step": 94310 + }, + { + "epoch": 27.442537096304918, + "grad_norm": 0.13149873912334442, + "learning_rate": 3.6022369584427755e-06, + "loss": 0.013, + "step": 94320 + }, + { + "epoch": 27.44544661041606, + "grad_norm": 0.0947219654917717, + "learning_rate": 3.5995501001106234e-06, + "loss": 0.0111, + "step": 94330 + }, + { + "epoch": 27.448356124527205, + "grad_norm": 0.08003505319356918, + "learning_rate": 3.5968641075674068e-06, + "loss": 0.0103, + "step": 94340 + }, + { + "epoch": 27.451265638638347, + "grad_norm": 0.10896062850952148, + "learning_rate": 3.5941789810171043e-06, + "loss": 0.0102, + "step": 94350 + }, + { + "epoch": 27.454175152749492, + "grad_norm": 0.06824241578578949, + "learning_rate": 3.591494720663636e-06, + "loss": 0.0102, + "step": 94360 + }, + { + "epoch": 27.457084666860634, + "grad_norm": 0.08297097682952881, + "learning_rate": 3.588811326710859e-06, + "loss": 0.0104, + "step": 94370 + }, + { + "epoch": 27.459994180971776, + "grad_norm": 0.12354122847318649, + "learning_rate": 3.5861287993625536e-06, + "loss": 0.0112, + "step": 94380 + }, + { + "epoch": 27.46290369508292, + "grad_norm": 0.18749909102916718, + "learning_rate": 3.5834471388224334e-06, + "loss": 0.0122, + "step": 94390 + }, + { + "epoch": 27.465813209194064, + "grad_norm": 0.11915509402751923, + "learning_rate": 3.5807663452941685e-06, + "loss": 0.0114, + "step": 94400 + }, + { + "epoch": 27.46872272330521, + "grad_norm": 0.11411920189857483, + "learning_rate": 3.5780864189813382e-06, + "loss": 0.0092, + "step": 94410 + }, + { + "epoch": 27.47163223741635, + "grad_norm": 0.3275917172431946, + "learning_rate": 3.575407360087463e-06, + "loss": 0.0129, + "step": 94420 + }, + { + "epoch": 27.474541751527497, + "grad_norm": 0.08812204003334045, + "learning_rate": 3.572729168816002e-06, + "loss": 0.0108, + "step": 94430 + }, + { + "epoch": 27.47745126563864, + "grad_norm": 0.10042457282543182, + "learning_rate": 3.5700518453703533e-06, + "loss": 0.0128, + "step": 94440 + }, + { + "epoch": 27.48036077974978, + "grad_norm": 0.1212499663233757, + "learning_rate": 3.5673753899538352e-06, + "loss": 0.0116, + "step": 94450 + }, + { + "epoch": 27.483270293860926, + "grad_norm": 0.1081278845667839, + "learning_rate": 3.5646998027697012e-06, + "loss": 0.01, + "step": 94460 + }, + { + "epoch": 27.486179807972068, + "grad_norm": 0.09333217144012451, + "learning_rate": 3.562025084021153e-06, + "loss": 0.011, + "step": 94470 + }, + { + "epoch": 27.489089322083213, + "grad_norm": 0.11088661849498749, + "learning_rate": 3.55935123391132e-06, + "loss": 0.0126, + "step": 94480 + }, + { + "epoch": 27.491998836194355, + "grad_norm": 0.12610630691051483, + "learning_rate": 3.556678252643254e-06, + "loss": 0.0139, + "step": 94490 + }, + { + "epoch": 27.4949083503055, + "grad_norm": 0.137221559882164, + "learning_rate": 3.5540061404199543e-06, + "loss": 0.0117, + "step": 94500 + }, + { + "epoch": 27.497817864416643, + "grad_norm": 0.10144466906785965, + "learning_rate": 3.5513348974443565e-06, + "loss": 0.0108, + "step": 94510 + }, + { + "epoch": 27.500727378527785, + "grad_norm": 0.19792458415031433, + "learning_rate": 3.5486645239193206e-06, + "loss": 0.0109, + "step": 94520 + }, + { + "epoch": 27.50363689263893, + "grad_norm": 0.10946356505155563, + "learning_rate": 3.5459950200476356e-06, + "loss": 0.0106, + "step": 94530 + }, + { + "epoch": 27.506546406750072, + "grad_norm": 0.10513915866613388, + "learning_rate": 3.5433263860320393e-06, + "loss": 0.0112, + "step": 94540 + }, + { + "epoch": 27.509455920861217, + "grad_norm": 0.11364459246397018, + "learning_rate": 3.5406586220751995e-06, + "loss": 0.0105, + "step": 94550 + }, + { + "epoch": 27.51236543497236, + "grad_norm": 0.06953130662441254, + "learning_rate": 3.53799172837971e-06, + "loss": 0.01, + "step": 94560 + }, + { + "epoch": 27.5152749490835, + "grad_norm": 0.10619767010211945, + "learning_rate": 3.535325705148109e-06, + "loss": 0.0109, + "step": 94570 + }, + { + "epoch": 27.518184463194647, + "grad_norm": 0.11646179109811783, + "learning_rate": 3.5326605525828558e-06, + "loss": 0.0106, + "step": 94580 + }, + { + "epoch": 27.52109397730579, + "grad_norm": 0.08058073371648788, + "learning_rate": 3.5299962708863563e-06, + "loss": 0.0103, + "step": 94590 + }, + { + "epoch": 27.524003491416934, + "grad_norm": 0.09547862410545349, + "learning_rate": 3.5273328602609504e-06, + "loss": 0.0115, + "step": 94600 + }, + { + "epoch": 27.526913005528076, + "grad_norm": 0.09992432594299316, + "learning_rate": 3.524670320908895e-06, + "loss": 0.011, + "step": 94610 + }, + { + "epoch": 27.52982251963922, + "grad_norm": 0.12963585555553436, + "learning_rate": 3.5220086530324027e-06, + "loss": 0.0105, + "step": 94620 + }, + { + "epoch": 27.532732033750364, + "grad_norm": 0.1298833042383194, + "learning_rate": 3.5193478568336e-06, + "loss": 0.0125, + "step": 94630 + }, + { + "epoch": 27.535641547861506, + "grad_norm": 0.11170578002929688, + "learning_rate": 3.516687932514568e-06, + "loss": 0.0114, + "step": 94640 + }, + { + "epoch": 27.53855106197265, + "grad_norm": 0.1346655786037445, + "learning_rate": 3.5140288802772973e-06, + "loss": 0.0109, + "step": 94650 + }, + { + "epoch": 27.541460576083793, + "grad_norm": 0.08231014013290405, + "learning_rate": 3.5113707003237355e-06, + "loss": 0.0106, + "step": 94660 + }, + { + "epoch": 27.54437009019494, + "grad_norm": 0.10566341131925583, + "learning_rate": 3.5087133928557526e-06, + "loss": 0.0102, + "step": 94670 + }, + { + "epoch": 27.54727960430608, + "grad_norm": 0.12202590703964233, + "learning_rate": 3.506056958075152e-06, + "loss": 0.0108, + "step": 94680 + }, + { + "epoch": 27.550189118417226, + "grad_norm": 0.11222313344478607, + "learning_rate": 3.5034013961836687e-06, + "loss": 0.0138, + "step": 94690 + }, + { + "epoch": 27.553098632528368, + "grad_norm": 0.2648308277130127, + "learning_rate": 3.500746707382977e-06, + "loss": 0.0115, + "step": 94700 + }, + { + "epoch": 27.55600814663951, + "grad_norm": 0.11764099448919296, + "learning_rate": 3.49809289187469e-06, + "loss": 0.0107, + "step": 94710 + }, + { + "epoch": 27.558917660750655, + "grad_norm": 0.09712879359722137, + "learning_rate": 3.4954399498603373e-06, + "loss": 0.01, + "step": 94720 + }, + { + "epoch": 27.561827174861797, + "grad_norm": 0.08830895274877548, + "learning_rate": 3.4927878815413953e-06, + "loss": 0.0106, + "step": 94730 + }, + { + "epoch": 27.564736688972943, + "grad_norm": 0.10168620944023132, + "learning_rate": 3.4901366871192784e-06, + "loss": 0.0117, + "step": 94740 + }, + { + "epoch": 27.567646203084085, + "grad_norm": 0.09465060383081436, + "learning_rate": 3.4874863667953224e-06, + "loss": 0.0099, + "step": 94750 + }, + { + "epoch": 27.57055571719523, + "grad_norm": 0.13587167859077454, + "learning_rate": 3.4848369207707955e-06, + "loss": 0.0121, + "step": 94760 + }, + { + "epoch": 27.573465231306372, + "grad_norm": 0.1235727071762085, + "learning_rate": 3.482188349246912e-06, + "loss": 0.0126, + "step": 94770 + }, + { + "epoch": 27.576374745417514, + "grad_norm": 0.1250971257686615, + "learning_rate": 3.4795406524248162e-06, + "loss": 0.011, + "step": 94780 + }, + { + "epoch": 27.57928425952866, + "grad_norm": 0.11103999614715576, + "learning_rate": 3.476893830505579e-06, + "loss": 0.0111, + "step": 94790 + }, + { + "epoch": 27.5821937736398, + "grad_norm": 0.12277573347091675, + "learning_rate": 3.4742478836902013e-06, + "loss": 0.012, + "step": 94800 + }, + { + "epoch": 27.585103287750947, + "grad_norm": 0.1308079957962036, + "learning_rate": 3.4716028121796404e-06, + "loss": 0.0117, + "step": 94810 + }, + { + "epoch": 27.58801280186209, + "grad_norm": 0.08796503394842148, + "learning_rate": 3.468958616174768e-06, + "loss": 0.0098, + "step": 94820 + }, + { + "epoch": 27.59092231597323, + "grad_norm": 0.10759902000427246, + "learning_rate": 3.4663152958763838e-06, + "loss": 0.0094, + "step": 94830 + }, + { + "epoch": 27.593831830084376, + "grad_norm": 0.15412397682666779, + "learning_rate": 3.4636728514852383e-06, + "loss": 0.0099, + "step": 94840 + }, + { + "epoch": 27.596741344195518, + "grad_norm": 0.13122279942035675, + "learning_rate": 3.4610312832020106e-06, + "loss": 0.01, + "step": 94850 + }, + { + "epoch": 27.599650858306664, + "grad_norm": 0.09805316478013992, + "learning_rate": 3.458390591227302e-06, + "loss": 0.0102, + "step": 94860 + }, + { + "epoch": 27.602560372417805, + "grad_norm": 0.07100105285644531, + "learning_rate": 3.455750775761666e-06, + "loss": 0.0113, + "step": 94870 + }, + { + "epoch": 27.60546988652895, + "grad_norm": 0.10157051682472229, + "learning_rate": 3.453111837005566e-06, + "loss": 0.0126, + "step": 94880 + }, + { + "epoch": 27.608379400640093, + "grad_norm": 0.10816706717014313, + "learning_rate": 3.4504737751594257e-06, + "loss": 0.0106, + "step": 94890 + }, + { + "epoch": 27.611288914751235, + "grad_norm": 0.11556370556354523, + "learning_rate": 3.4478365904235764e-06, + "loss": 0.0116, + "step": 94900 + }, + { + "epoch": 27.61419842886238, + "grad_norm": 0.17917726933956146, + "learning_rate": 3.4452002829983016e-06, + "loss": 0.0129, + "step": 94910 + }, + { + "epoch": 27.617107942973522, + "grad_norm": 0.06500552594661713, + "learning_rate": 3.4425648530838124e-06, + "loss": 0.0092, + "step": 94920 + }, + { + "epoch": 27.620017457084668, + "grad_norm": 0.10232625156641006, + "learning_rate": 3.4399303008802486e-06, + "loss": 0.0114, + "step": 94930 + }, + { + "epoch": 27.62292697119581, + "grad_norm": 0.08868725597858429, + "learning_rate": 3.4372966265876905e-06, + "loss": 0.0106, + "step": 94940 + }, + { + "epoch": 27.625836485306955, + "grad_norm": 0.12720325589179993, + "learning_rate": 3.4346638304061428e-06, + "loss": 0.0117, + "step": 94950 + }, + { + "epoch": 27.628745999418097, + "grad_norm": 0.07546389847993851, + "learning_rate": 3.4320319125355544e-06, + "loss": 0.0104, + "step": 94960 + }, + { + "epoch": 27.63165551352924, + "grad_norm": 0.10670248419046402, + "learning_rate": 3.429400873175803e-06, + "loss": 0.0119, + "step": 94970 + }, + { + "epoch": 27.634565027640384, + "grad_norm": 0.11871220916509628, + "learning_rate": 3.426770712526696e-06, + "loss": 0.0115, + "step": 94980 + }, + { + "epoch": 27.637474541751526, + "grad_norm": 0.1486886739730835, + "learning_rate": 3.424141430787972e-06, + "loss": 0.0119, + "step": 94990 + }, + { + "epoch": 27.640384055862672, + "grad_norm": 0.10083924233913422, + "learning_rate": 3.4215130281593125e-06, + "loss": 0.0115, + "step": 95000 + }, + { + "epoch": 27.643293569973814, + "grad_norm": 0.10269910842180252, + "learning_rate": 3.4188855048403316e-06, + "loss": 0.0104, + "step": 95010 + }, + { + "epoch": 27.64620308408496, + "grad_norm": 0.11101341992616653, + "learning_rate": 3.4162588610305643e-06, + "loss": 0.0086, + "step": 95020 + }, + { + "epoch": 27.6491125981961, + "grad_norm": 0.07598558068275452, + "learning_rate": 3.41363309692949e-06, + "loss": 0.0131, + "step": 95030 + }, + { + "epoch": 27.652022112307243, + "grad_norm": 0.1205715462565422, + "learning_rate": 3.4110082127365233e-06, + "loss": 0.0126, + "step": 95040 + }, + { + "epoch": 27.65493162641839, + "grad_norm": 0.12604796886444092, + "learning_rate": 3.408384208651002e-06, + "loss": 0.0103, + "step": 95050 + }, + { + "epoch": 27.65784114052953, + "grad_norm": 0.11156399548053741, + "learning_rate": 3.4057610848721965e-06, + "loss": 0.0115, + "step": 95060 + }, + { + "epoch": 27.660750654640676, + "grad_norm": 0.08702186495065689, + "learning_rate": 3.4031388415993227e-06, + "loss": 0.0104, + "step": 95070 + }, + { + "epoch": 27.663660168751818, + "grad_norm": 0.06884746998548508, + "learning_rate": 3.400517479031524e-06, + "loss": 0.0123, + "step": 95080 + }, + { + "epoch": 27.666569682862963, + "grad_norm": 0.13732360303401947, + "learning_rate": 3.3978969973678754e-06, + "loss": 0.0111, + "step": 95090 + }, + { + "epoch": 27.669479196974105, + "grad_norm": 0.12410221248865128, + "learning_rate": 3.395277396807372e-06, + "loss": 0.0135, + "step": 95100 + }, + { + "epoch": 27.672388711085247, + "grad_norm": 0.11395318806171417, + "learning_rate": 3.392658677548975e-06, + "loss": 0.0118, + "step": 95110 + }, + { + "epoch": 27.675298225196393, + "grad_norm": 0.16077199578285217, + "learning_rate": 3.39004083979155e-06, + "loss": 0.01, + "step": 95120 + }, + { + "epoch": 27.678207739307535, + "grad_norm": 0.11561659723520279, + "learning_rate": 3.3874238837339e-06, + "loss": 0.0107, + "step": 95130 + }, + { + "epoch": 27.68111725341868, + "grad_norm": 0.09501642733812332, + "learning_rate": 3.384807809574769e-06, + "loss": 0.01, + "step": 95140 + }, + { + "epoch": 27.684026767529822, + "grad_norm": 0.11177954077720642, + "learning_rate": 3.3821926175128355e-06, + "loss": 0.0123, + "step": 95150 + }, + { + "epoch": 27.686936281640968, + "grad_norm": 0.07810449600219727, + "learning_rate": 3.3795783077467015e-06, + "loss": 0.0117, + "step": 95160 + }, + { + "epoch": 27.68984579575211, + "grad_norm": 0.1867227703332901, + "learning_rate": 3.376964880474903e-06, + "loss": 0.0121, + "step": 95170 + }, + { + "epoch": 27.69275530986325, + "grad_norm": 0.09648644924163818, + "learning_rate": 3.374352335895917e-06, + "loss": 0.0102, + "step": 95180 + }, + { + "epoch": 27.695664823974397, + "grad_norm": 0.11608356237411499, + "learning_rate": 3.371740674208153e-06, + "loss": 0.01, + "step": 95190 + }, + { + "epoch": 27.69857433808554, + "grad_norm": 0.09437508136034012, + "learning_rate": 3.36912989560994e-06, + "loss": 0.012, + "step": 95200 + }, + { + "epoch": 27.701483852196684, + "grad_norm": 0.12725184857845306, + "learning_rate": 3.366520000299554e-06, + "loss": 0.0109, + "step": 95210 + }, + { + "epoch": 27.704393366307826, + "grad_norm": 0.11633815616369247, + "learning_rate": 3.363910988475205e-06, + "loss": 0.0108, + "step": 95220 + }, + { + "epoch": 27.707302880418972, + "grad_norm": 0.06640446186065674, + "learning_rate": 3.3613028603350203e-06, + "loss": 0.0122, + "step": 95230 + }, + { + "epoch": 27.710212394530114, + "grad_norm": 0.11604632437229156, + "learning_rate": 3.358695616077078e-06, + "loss": 0.011, + "step": 95240 + }, + { + "epoch": 27.713121908641256, + "grad_norm": 0.12904050946235657, + "learning_rate": 3.356089255899375e-06, + "loss": 0.0124, + "step": 95250 + }, + { + "epoch": 27.7160314227524, + "grad_norm": 0.11814986914396286, + "learning_rate": 3.3534837799998524e-06, + "loss": 0.0103, + "step": 95260 + }, + { + "epoch": 27.718940936863543, + "grad_norm": 0.09813809394836426, + "learning_rate": 3.350879188576374e-06, + "loss": 0.0118, + "step": 95270 + }, + { + "epoch": 27.72185045097469, + "grad_norm": 0.10894063115119934, + "learning_rate": 3.348275481826746e-06, + "loss": 0.0116, + "step": 95280 + }, + { + "epoch": 27.72475996508583, + "grad_norm": 0.09469447284936905, + "learning_rate": 3.345672659948698e-06, + "loss": 0.0109, + "step": 95290 + }, + { + "epoch": 27.727669479196972, + "grad_norm": 0.14661753177642822, + "learning_rate": 3.3430707231398994e-06, + "loss": 0.0148, + "step": 95300 + }, + { + "epoch": 27.730578993308118, + "grad_norm": 0.1303730309009552, + "learning_rate": 3.340469671597954e-06, + "loss": 0.0106, + "step": 95310 + }, + { + "epoch": 27.73348850741926, + "grad_norm": 0.09840922057628632, + "learning_rate": 3.3378695055203866e-06, + "loss": 0.0108, + "step": 95320 + }, + { + "epoch": 27.736398021530405, + "grad_norm": 0.0980444848537445, + "learning_rate": 3.335270225104667e-06, + "loss": 0.0109, + "step": 95330 + }, + { + "epoch": 27.739307535641547, + "grad_norm": 0.13783204555511475, + "learning_rate": 3.332671830548198e-06, + "loss": 0.0118, + "step": 95340 + }, + { + "epoch": 27.742217049752693, + "grad_norm": 0.14507131278514862, + "learning_rate": 3.3300743220483037e-06, + "loss": 0.0124, + "step": 95350 + }, + { + "epoch": 27.745126563863835, + "grad_norm": 0.08336365967988968, + "learning_rate": 3.327477699802248e-06, + "loss": 0.0093, + "step": 95360 + }, + { + "epoch": 27.748036077974977, + "grad_norm": 0.13848307728767395, + "learning_rate": 3.3248819640072285e-06, + "loss": 0.0106, + "step": 95370 + }, + { + "epoch": 27.750945592086122, + "grad_norm": 0.16974091529846191, + "learning_rate": 3.322287114860379e-06, + "loss": 0.0133, + "step": 95380 + }, + { + "epoch": 27.753855106197264, + "grad_norm": 0.09915503114461899, + "learning_rate": 3.3196931525587566e-06, + "loss": 0.0101, + "step": 95390 + }, + { + "epoch": 27.75676462030841, + "grad_norm": 0.12499987334012985, + "learning_rate": 3.317100077299347e-06, + "loss": 0.0101, + "step": 95400 + }, + { + "epoch": 27.75967413441955, + "grad_norm": 0.08609102666378021, + "learning_rate": 3.3145078892790947e-06, + "loss": 0.0114, + "step": 95410 + }, + { + "epoch": 27.762583648530697, + "grad_norm": 0.14108934998512268, + "learning_rate": 3.311916588694852e-06, + "loss": 0.0103, + "step": 95420 + }, + { + "epoch": 27.76549316264184, + "grad_norm": 0.11490854620933533, + "learning_rate": 3.3093261757434035e-06, + "loss": 0.0099, + "step": 95430 + }, + { + "epoch": 27.76840267675298, + "grad_norm": 0.09430652856826782, + "learning_rate": 3.3067366506214814e-06, + "loss": 0.0107, + "step": 95440 + }, + { + "epoch": 27.771312190864126, + "grad_norm": 0.08889011293649673, + "learning_rate": 3.304148013525747e-06, + "loss": 0.0114, + "step": 95450 + }, + { + "epoch": 27.774221704975268, + "grad_norm": 0.09472872316837311, + "learning_rate": 3.3015602646527843e-06, + "loss": 0.0109, + "step": 95460 + }, + { + "epoch": 27.777131219086414, + "grad_norm": 0.06845200061798096, + "learning_rate": 3.2989734041991114e-06, + "loss": 0.0108, + "step": 95470 + }, + { + "epoch": 27.780040733197556, + "grad_norm": 0.14440305531024933, + "learning_rate": 3.2963874323611897e-06, + "loss": 0.0111, + "step": 95480 + }, + { + "epoch": 27.7829502473087, + "grad_norm": 0.13902699947357178, + "learning_rate": 3.2938023493354086e-06, + "loss": 0.0121, + "step": 95490 + }, + { + "epoch": 27.785859761419843, + "grad_norm": 0.06112741306424141, + "learning_rate": 3.291218155318081e-06, + "loss": 0.0097, + "step": 95500 + }, + { + "epoch": 27.788769275530985, + "grad_norm": 0.10871411859989166, + "learning_rate": 3.2886348505054633e-06, + "loss": 0.0114, + "step": 95510 + }, + { + "epoch": 27.79167878964213, + "grad_norm": 0.1257927268743515, + "learning_rate": 3.286052435093745e-06, + "loss": 0.0117, + "step": 95520 + }, + { + "epoch": 27.794588303753272, + "grad_norm": 0.15815994143486023, + "learning_rate": 3.283470909279038e-06, + "loss": 0.0129, + "step": 95530 + }, + { + "epoch": 27.797497817864418, + "grad_norm": 0.0984252318739891, + "learning_rate": 3.2808902732573896e-06, + "loss": 0.0131, + "step": 95540 + }, + { + "epoch": 27.80040733197556, + "grad_norm": 0.10689353942871094, + "learning_rate": 3.278310527224784e-06, + "loss": 0.0114, + "step": 95550 + }, + { + "epoch": 27.8033168460867, + "grad_norm": 0.05862057954072952, + "learning_rate": 3.2757316713771444e-06, + "loss": 0.0095, + "step": 95560 + }, + { + "epoch": 27.806226360197847, + "grad_norm": 0.1461976319551468, + "learning_rate": 3.2731537059103054e-06, + "loss": 0.0092, + "step": 95570 + }, + { + "epoch": 27.80913587430899, + "grad_norm": 0.10165470093488693, + "learning_rate": 3.2705766310200557e-06, + "loss": 0.0111, + "step": 95580 + }, + { + "epoch": 27.812045388420135, + "grad_norm": 0.20507851243019104, + "learning_rate": 3.2680004469021003e-06, + "loss": 0.0102, + "step": 95590 + }, + { + "epoch": 27.814954902531277, + "grad_norm": 0.08794418722391129, + "learning_rate": 3.265425153752086e-06, + "loss": 0.0114, + "step": 95600 + }, + { + "epoch": 27.817864416642422, + "grad_norm": 0.09608937054872513, + "learning_rate": 3.2628507517655927e-06, + "loss": 0.0115, + "step": 95610 + }, + { + "epoch": 27.820773930753564, + "grad_norm": 0.07978928089141846, + "learning_rate": 3.2602772411381243e-06, + "loss": 0.0132, + "step": 95620 + }, + { + "epoch": 27.823683444864706, + "grad_norm": 0.08332649618387222, + "learning_rate": 3.2577046220651264e-06, + "loss": 0.0113, + "step": 95630 + }, + { + "epoch": 27.82659295897585, + "grad_norm": 0.08877512067556381, + "learning_rate": 3.2551328947419675e-06, + "loss": 0.0114, + "step": 95640 + }, + { + "epoch": 27.829502473086993, + "grad_norm": 0.13399377465248108, + "learning_rate": 3.252562059363959e-06, + "loss": 0.0134, + "step": 95650 + }, + { + "epoch": 27.83241198719814, + "grad_norm": 0.14602626860141754, + "learning_rate": 3.249992116126331e-06, + "loss": 0.0126, + "step": 95660 + }, + { + "epoch": 27.83532150130928, + "grad_norm": 0.09860152006149292, + "learning_rate": 3.24742306522426e-06, + "loss": 0.0114, + "step": 95670 + }, + { + "epoch": 27.838231015420426, + "grad_norm": 0.07603809982538223, + "learning_rate": 3.2448549068528505e-06, + "loss": 0.0117, + "step": 95680 + }, + { + "epoch": 27.841140529531568, + "grad_norm": 0.14143702387809753, + "learning_rate": 3.2422876412071324e-06, + "loss": 0.0122, + "step": 95690 + }, + { + "epoch": 27.84405004364271, + "grad_norm": 0.08508665859699249, + "learning_rate": 3.2397212684820665e-06, + "loss": 0.0108, + "step": 95700 + }, + { + "epoch": 27.846959557753856, + "grad_norm": 0.3370261788368225, + "learning_rate": 3.237155788872569e-06, + "loss": 0.0098, + "step": 95710 + }, + { + "epoch": 27.849869071864997, + "grad_norm": 0.13603205978870392, + "learning_rate": 3.2345912025734586e-06, + "loss": 0.0112, + "step": 95720 + }, + { + "epoch": 27.852778585976143, + "grad_norm": 0.10424172878265381, + "learning_rate": 3.2320275097794997e-06, + "loss": 0.0118, + "step": 95730 + }, + { + "epoch": 27.855688100087285, + "grad_norm": 0.09976571053266525, + "learning_rate": 3.229464710685389e-06, + "loss": 0.0101, + "step": 95740 + }, + { + "epoch": 27.85859761419843, + "grad_norm": 0.14630518853664398, + "learning_rate": 3.2269028054857576e-06, + "loss": 0.0117, + "step": 95750 + }, + { + "epoch": 27.861507128309572, + "grad_norm": 0.08400361984968185, + "learning_rate": 3.2243417943751623e-06, + "loss": 0.011, + "step": 95760 + }, + { + "epoch": 27.864416642420714, + "grad_norm": 0.10217839479446411, + "learning_rate": 3.221781677548093e-06, + "loss": 0.0117, + "step": 95770 + }, + { + "epoch": 27.86732615653186, + "grad_norm": 0.1014319434762001, + "learning_rate": 3.2192224551989734e-06, + "loss": 0.011, + "step": 95780 + }, + { + "epoch": 27.870235670643, + "grad_norm": 0.10091187059879303, + "learning_rate": 3.2166641275221688e-06, + "loss": 0.012, + "step": 95790 + }, + { + "epoch": 27.873145184754147, + "grad_norm": 0.074745312333107, + "learning_rate": 3.2141066947119547e-06, + "loss": 0.0105, + "step": 95800 + }, + { + "epoch": 27.87605469886529, + "grad_norm": 0.11889063566923141, + "learning_rate": 3.211550156962556e-06, + "loss": 0.011, + "step": 95810 + }, + { + "epoch": 27.878964212976435, + "grad_norm": 0.12420188635587692, + "learning_rate": 3.208994514468131e-06, + "loss": 0.0095, + "step": 95820 + }, + { + "epoch": 27.881873727087576, + "grad_norm": 0.09920429438352585, + "learning_rate": 3.206439767422759e-06, + "loss": 0.0116, + "step": 95830 + }, + { + "epoch": 27.88478324119872, + "grad_norm": 0.1052580177783966, + "learning_rate": 3.203885916020452e-06, + "loss": 0.0122, + "step": 95840 + }, + { + "epoch": 27.887692755309864, + "grad_norm": 0.112399160861969, + "learning_rate": 3.2013329604551628e-06, + "loss": 0.0133, + "step": 95850 + }, + { + "epoch": 27.890602269421006, + "grad_norm": 0.0980461910367012, + "learning_rate": 3.198780900920773e-06, + "loss": 0.0132, + "step": 95860 + }, + { + "epoch": 27.89351178353215, + "grad_norm": 0.08061207085847855, + "learning_rate": 3.1962297376110896e-06, + "loss": 0.0125, + "step": 95870 + }, + { + "epoch": 27.896421297643293, + "grad_norm": 0.07072219997644424, + "learning_rate": 3.1936794707198635e-06, + "loss": 0.0095, + "step": 95880 + }, + { + "epoch": 27.89933081175444, + "grad_norm": 0.09454259276390076, + "learning_rate": 3.1911301004407622e-06, + "loss": 0.0108, + "step": 95890 + }, + { + "epoch": 27.90224032586558, + "grad_norm": 0.07414795458316803, + "learning_rate": 3.188581626967402e-06, + "loss": 0.0121, + "step": 95900 + }, + { + "epoch": 27.905149839976723, + "grad_norm": 0.11219244450330734, + "learning_rate": 3.1860340504933155e-06, + "loss": 0.0096, + "step": 95910 + }, + { + "epoch": 27.908059354087868, + "grad_norm": 0.1459604650735855, + "learning_rate": 3.1834873712119773e-06, + "loss": 0.0114, + "step": 95920 + }, + { + "epoch": 27.91096886819901, + "grad_norm": 0.09956956654787064, + "learning_rate": 3.1809415893167944e-06, + "loss": 0.0108, + "step": 95930 + }, + { + "epoch": 27.913878382310156, + "grad_norm": 0.12410473823547363, + "learning_rate": 3.1783967050010954e-06, + "loss": 0.012, + "step": 95940 + }, + { + "epoch": 27.916787896421297, + "grad_norm": 0.09649443626403809, + "learning_rate": 3.1758527184581544e-06, + "loss": 0.0121, + "step": 95950 + }, + { + "epoch": 27.91969741053244, + "grad_norm": 0.10246909409761429, + "learning_rate": 3.173309629881163e-06, + "loss": 0.0131, + "step": 95960 + }, + { + "epoch": 27.922606924643585, + "grad_norm": 0.10874198377132416, + "learning_rate": 3.170767439463256e-06, + "loss": 0.0099, + "step": 95970 + }, + { + "epoch": 27.925516438754727, + "grad_norm": 0.10506575554609299, + "learning_rate": 3.1682261473974995e-06, + "loss": 0.0118, + "step": 95980 + }, + { + "epoch": 27.928425952865872, + "grad_norm": 0.09434299916028976, + "learning_rate": 3.165685753876885e-06, + "loss": 0.0117, + "step": 95990 + }, + { + "epoch": 27.931335466977014, + "grad_norm": 0.15947142243385315, + "learning_rate": 3.1631462590943337e-06, + "loss": 0.011, + "step": 96000 + }, + { + "epoch": 27.93424498108816, + "grad_norm": 0.12598560750484467, + "learning_rate": 3.160607663242706e-06, + "loss": 0.0101, + "step": 96010 + }, + { + "epoch": 27.9371544951993, + "grad_norm": 0.0969325453042984, + "learning_rate": 3.1580699665147987e-06, + "loss": 0.0121, + "step": 96020 + }, + { + "epoch": 27.940064009310444, + "grad_norm": 0.09397216886281967, + "learning_rate": 3.155533169103323e-06, + "loss": 0.011, + "step": 96030 + }, + { + "epoch": 27.94297352342159, + "grad_norm": 0.12951764464378357, + "learning_rate": 3.1529972712009358e-06, + "loss": 0.0094, + "step": 96040 + }, + { + "epoch": 27.94588303753273, + "grad_norm": 0.10417152941226959, + "learning_rate": 3.1504622730002287e-06, + "loss": 0.0108, + "step": 96050 + }, + { + "epoch": 27.948792551643876, + "grad_norm": 0.09746059030294418, + "learning_rate": 3.147928174693709e-06, + "loss": 0.0113, + "step": 96060 + }, + { + "epoch": 27.95170206575502, + "grad_norm": 0.07501856982707977, + "learning_rate": 3.145394976473826e-06, + "loss": 0.0117, + "step": 96070 + }, + { + "epoch": 27.954611579866164, + "grad_norm": 0.16481079161167145, + "learning_rate": 3.1428626785329594e-06, + "loss": 0.0096, + "step": 96080 + }, + { + "epoch": 27.957521093977306, + "grad_norm": 0.22736619412899017, + "learning_rate": 3.1403312810634255e-06, + "loss": 0.0108, + "step": 96090 + }, + { + "epoch": 27.960430608088448, + "grad_norm": 0.10124709457159042, + "learning_rate": 3.137800784257465e-06, + "loss": 0.0111, + "step": 96100 + }, + { + "epoch": 27.963340122199593, + "grad_norm": 0.1162559986114502, + "learning_rate": 3.1352711883072434e-06, + "loss": 0.0099, + "step": 96110 + }, + { + "epoch": 27.966249636310735, + "grad_norm": 0.08285607397556305, + "learning_rate": 3.132742493404881e-06, + "loss": 0.0093, + "step": 96120 + }, + { + "epoch": 27.96915915042188, + "grad_norm": 0.23184438049793243, + "learning_rate": 3.1302146997424114e-06, + "loss": 0.0102, + "step": 96130 + }, + { + "epoch": 27.972068664533023, + "grad_norm": 0.15531349182128906, + "learning_rate": 3.127687807511796e-06, + "loss": 0.0092, + "step": 96140 + }, + { + "epoch": 27.974978178644168, + "grad_norm": 0.09367205202579498, + "learning_rate": 3.1251618169049413e-06, + "loss": 0.0091, + "step": 96150 + }, + { + "epoch": 27.97788769275531, + "grad_norm": 0.0954299196600914, + "learning_rate": 3.1226367281136835e-06, + "loss": 0.0105, + "step": 96160 + }, + { + "epoch": 27.980797206866452, + "grad_norm": 0.08079216629266739, + "learning_rate": 3.12011254132978e-06, + "loss": 0.0107, + "step": 96170 + }, + { + "epoch": 27.983706720977597, + "grad_norm": 0.13292697072029114, + "learning_rate": 3.1175892567449316e-06, + "loss": 0.0107, + "step": 96180 + }, + { + "epoch": 27.98661623508874, + "grad_norm": 0.11988982558250427, + "learning_rate": 3.1150668745507573e-06, + "loss": 0.0104, + "step": 96190 + }, + { + "epoch": 27.989525749199885, + "grad_norm": 0.0814858004450798, + "learning_rate": 3.112545394938825e-06, + "loss": 0.0106, + "step": 96200 + }, + { + "epoch": 27.992435263311027, + "grad_norm": 0.20008736848831177, + "learning_rate": 3.110024818100615e-06, + "loss": 0.0106, + "step": 96210 + }, + { + "epoch": 27.995344777422172, + "grad_norm": 0.10253681242465973, + "learning_rate": 3.107505144227553e-06, + "loss": 0.0119, + "step": 96220 + }, + { + "epoch": 27.998254291533314, + "grad_norm": 0.12773674726486206, + "learning_rate": 3.104986373510996e-06, + "loss": 0.0095, + "step": 96230 + }, + { + "epoch": 28.001163805644456, + "grad_norm": 0.15294595062732697, + "learning_rate": 3.1024685061422196e-06, + "loss": 0.0093, + "step": 96240 + }, + { + "epoch": 28.0040733197556, + "grad_norm": 0.1709751933813095, + "learning_rate": 3.0999515423124473e-06, + "loss": 0.0118, + "step": 96250 + }, + { + "epoch": 28.006982833866743, + "grad_norm": 0.18763971328735352, + "learning_rate": 3.0974354822128174e-06, + "loss": 0.011, + "step": 96260 + }, + { + "epoch": 28.00989234797789, + "grad_norm": 0.0964696928858757, + "learning_rate": 3.094920326034417e-06, + "loss": 0.011, + "step": 96270 + }, + { + "epoch": 28.01280186208903, + "grad_norm": 0.09700702130794525, + "learning_rate": 3.092406073968247e-06, + "loss": 0.0123, + "step": 96280 + }, + { + "epoch": 28.015711376200173, + "grad_norm": 0.08651600778102875, + "learning_rate": 3.089892726205257e-06, + "loss": 0.0101, + "step": 96290 + }, + { + "epoch": 28.01862089031132, + "grad_norm": 0.05795402079820633, + "learning_rate": 3.0873802829363108e-06, + "loss": 0.0107, + "step": 96300 + }, + { + "epoch": 28.02153040442246, + "grad_norm": 0.1959853172302246, + "learning_rate": 3.0848687443522145e-06, + "loss": 0.01, + "step": 96310 + }, + { + "epoch": 28.024439918533606, + "grad_norm": 0.10285027325153351, + "learning_rate": 3.0823581106437093e-06, + "loss": 0.0111, + "step": 96320 + }, + { + "epoch": 28.027349432644748, + "grad_norm": 0.10721846669912338, + "learning_rate": 3.079848382001451e-06, + "loss": 0.011, + "step": 96330 + }, + { + "epoch": 28.030258946755893, + "grad_norm": 0.09539555013179779, + "learning_rate": 3.0773395586160423e-06, + "loss": 0.0102, + "step": 96340 + }, + { + "epoch": 28.033168460867035, + "grad_norm": 0.09827952086925507, + "learning_rate": 3.074831640678014e-06, + "loss": 0.0088, + "step": 96350 + }, + { + "epoch": 28.036077974978177, + "grad_norm": 0.09395992755889893, + "learning_rate": 3.0723246283778256e-06, + "loss": 0.0102, + "step": 96360 + }, + { + "epoch": 28.038987489089322, + "grad_norm": 0.11089697480201721, + "learning_rate": 3.069818521905861e-06, + "loss": 0.0122, + "step": 96370 + }, + { + "epoch": 28.041897003200464, + "grad_norm": 0.06394951790571213, + "learning_rate": 3.0673133214524467e-06, + "loss": 0.0122, + "step": 96380 + }, + { + "epoch": 28.04480651731161, + "grad_norm": 0.10737595707178116, + "learning_rate": 3.0648090272078404e-06, + "loss": 0.0109, + "step": 96390 + }, + { + "epoch": 28.047716031422752, + "grad_norm": 0.09763885289430618, + "learning_rate": 3.062305639362225e-06, + "loss": 0.0091, + "step": 96400 + }, + { + "epoch": 28.050625545533897, + "grad_norm": 0.12139733880758286, + "learning_rate": 3.059803158105705e-06, + "loss": 0.0122, + "step": 96410 + }, + { + "epoch": 28.05353505964504, + "grad_norm": 0.14718031883239746, + "learning_rate": 3.0573015836283454e-06, + "loss": 0.0123, + "step": 96420 + }, + { + "epoch": 28.05644457375618, + "grad_norm": 0.13033831119537354, + "learning_rate": 3.054800916120115e-06, + "loss": 0.0115, + "step": 96430 + }, + { + "epoch": 28.059354087867327, + "grad_norm": 0.07716056704521179, + "learning_rate": 3.0523011557709195e-06, + "loss": 0.0121, + "step": 96440 + }, + { + "epoch": 28.06226360197847, + "grad_norm": 0.12140464782714844, + "learning_rate": 3.0498023027706036e-06, + "loss": 0.0123, + "step": 96450 + }, + { + "epoch": 28.065173116089614, + "grad_norm": 0.13308554887771606, + "learning_rate": 3.0473043573089414e-06, + "loss": 0.0118, + "step": 96460 + }, + { + "epoch": 28.068082630200756, + "grad_norm": 0.1049046739935875, + "learning_rate": 3.044807319575634e-06, + "loss": 0.0114, + "step": 96470 + }, + { + "epoch": 28.0709921443119, + "grad_norm": 0.0995122641324997, + "learning_rate": 3.0423111897603085e-06, + "loss": 0.0116, + "step": 96480 + }, + { + "epoch": 28.073901658423043, + "grad_norm": 0.08768175542354584, + "learning_rate": 3.039815968052534e-06, + "loss": 0.0093, + "step": 96490 + }, + { + "epoch": 28.076811172534185, + "grad_norm": 0.0760047435760498, + "learning_rate": 3.0373216546418105e-06, + "loss": 0.0101, + "step": 96500 + }, + { + "epoch": 28.07972068664533, + "grad_norm": 0.09915129840373993, + "learning_rate": 3.034828249717556e-06, + "loss": 0.0125, + "step": 96510 + }, + { + "epoch": 28.082630200756473, + "grad_norm": 0.1715427041053772, + "learning_rate": 3.032335753469135e-06, + "loss": 0.0094, + "step": 96520 + }, + { + "epoch": 28.08553971486762, + "grad_norm": 0.13833710551261902, + "learning_rate": 3.029844166085835e-06, + "loss": 0.0109, + "step": 96530 + }, + { + "epoch": 28.08844922897876, + "grad_norm": 0.07223451137542725, + "learning_rate": 3.027353487756872e-06, + "loss": 0.0115, + "step": 96540 + }, + { + "epoch": 28.091358743089906, + "grad_norm": 0.10387324541807175, + "learning_rate": 3.0248637186714027e-06, + "loss": 0.0117, + "step": 96550 + }, + { + "epoch": 28.094268257201048, + "grad_norm": 0.12927110493183136, + "learning_rate": 3.0223748590184997e-06, + "loss": 0.0108, + "step": 96560 + }, + { + "epoch": 28.09717777131219, + "grad_norm": 0.10737679898738861, + "learning_rate": 3.019886908987186e-06, + "loss": 0.0098, + "step": 96570 + }, + { + "epoch": 28.100087285423335, + "grad_norm": 0.09903975576162338, + "learning_rate": 3.0173998687663963e-06, + "loss": 0.012, + "step": 96580 + }, + { + "epoch": 28.102996799534477, + "grad_norm": 0.09906672686338425, + "learning_rate": 3.0149137385450104e-06, + "loss": 0.0101, + "step": 96590 + }, + { + "epoch": 28.105906313645622, + "grad_norm": 0.1351672261953354, + "learning_rate": 3.0124285185118295e-06, + "loss": 0.0115, + "step": 96600 + }, + { + "epoch": 28.108815827756764, + "grad_norm": 0.1094542145729065, + "learning_rate": 3.0099442088555906e-06, + "loss": 0.0113, + "step": 96610 + }, + { + "epoch": 28.11172534186791, + "grad_norm": 0.07726401090621948, + "learning_rate": 3.007460809764965e-06, + "loss": 0.0123, + "step": 96620 + }, + { + "epoch": 28.11463485597905, + "grad_norm": 0.14052297174930573, + "learning_rate": 3.004978321428544e-06, + "loss": 0.0115, + "step": 96630 + }, + { + "epoch": 28.117544370090194, + "grad_norm": 0.13830910623073578, + "learning_rate": 3.0024967440348585e-06, + "loss": 0.0145, + "step": 96640 + }, + { + "epoch": 28.12045388420134, + "grad_norm": 0.22034676373004913, + "learning_rate": 3.0000160777723728e-06, + "loss": 0.0096, + "step": 96650 + }, + { + "epoch": 28.12336339831248, + "grad_norm": 0.08394941687583923, + "learning_rate": 2.997536322829472e-06, + "loss": 0.0116, + "step": 96660 + }, + { + "epoch": 28.126272912423627, + "grad_norm": 0.07819589972496033, + "learning_rate": 2.995057479394476e-06, + "loss": 0.0088, + "step": 96670 + }, + { + "epoch": 28.12918242653477, + "grad_norm": 0.12791401147842407, + "learning_rate": 2.9925795476556378e-06, + "loss": 0.013, + "step": 96680 + }, + { + "epoch": 28.13209194064591, + "grad_norm": 0.16163764894008636, + "learning_rate": 2.9901025278011444e-06, + "loss": 0.0107, + "step": 96690 + }, + { + "epoch": 28.135001454757056, + "grad_norm": 0.1316872537136078, + "learning_rate": 2.987626420019107e-06, + "loss": 0.0107, + "step": 96700 + }, + { + "epoch": 28.137910968868198, + "grad_norm": 0.5938172340393066, + "learning_rate": 2.985151224497559e-06, + "loss": 0.0125, + "step": 96710 + }, + { + "epoch": 28.140820482979343, + "grad_norm": 0.18991714715957642, + "learning_rate": 2.982676941424492e-06, + "loss": 0.0127, + "step": 96720 + }, + { + "epoch": 28.143729997090485, + "grad_norm": 0.07440955936908722, + "learning_rate": 2.980203570987805e-06, + "loss": 0.0118, + "step": 96730 + }, + { + "epoch": 28.14663951120163, + "grad_norm": 0.09816870093345642, + "learning_rate": 2.977731113375329e-06, + "loss": 0.0106, + "step": 96740 + }, + { + "epoch": 28.149549025312773, + "grad_norm": 0.10693905502557755, + "learning_rate": 2.975259568774835e-06, + "loss": 0.0108, + "step": 96750 + }, + { + "epoch": 28.152458539423915, + "grad_norm": 0.20695947110652924, + "learning_rate": 2.972788937374024e-06, + "loss": 0.0122, + "step": 96760 + }, + { + "epoch": 28.15536805353506, + "grad_norm": 0.12052540481090546, + "learning_rate": 2.9703192193605194e-06, + "loss": 0.0126, + "step": 96770 + }, + { + "epoch": 28.158277567646202, + "grad_norm": 0.08253193646669388, + "learning_rate": 2.967850414921879e-06, + "loss": 0.0121, + "step": 96780 + }, + { + "epoch": 28.161187081757348, + "grad_norm": 0.08992686867713928, + "learning_rate": 2.9653825242455923e-06, + "loss": 0.0104, + "step": 96790 + }, + { + "epoch": 28.16409659586849, + "grad_norm": 0.09221556037664413, + "learning_rate": 2.9629155475190866e-06, + "loss": 0.0108, + "step": 96800 + }, + { + "epoch": 28.167006109979635, + "grad_norm": 0.08842688053846359, + "learning_rate": 2.9604494849297038e-06, + "loss": 0.0136, + "step": 96810 + }, + { + "epoch": 28.169915624090777, + "grad_norm": 0.09163209795951843, + "learning_rate": 2.9579843366647267e-06, + "loss": 0.0102, + "step": 96820 + }, + { + "epoch": 28.17282513820192, + "grad_norm": 0.12358071655035019, + "learning_rate": 2.955520102911375e-06, + "loss": 0.0121, + "step": 96830 + }, + { + "epoch": 28.175734652313064, + "grad_norm": 0.14167384803295135, + "learning_rate": 2.953056783856783e-06, + "loss": 0.0135, + "step": 96840 + }, + { + "epoch": 28.178644166424206, + "grad_norm": 0.11418069154024124, + "learning_rate": 2.950594379688023e-06, + "loss": 0.013, + "step": 96850 + }, + { + "epoch": 28.18155368053535, + "grad_norm": 0.06615874916315079, + "learning_rate": 2.948132890592099e-06, + "loss": 0.0092, + "step": 96860 + }, + { + "epoch": 28.184463194646494, + "grad_norm": 0.157173290848732, + "learning_rate": 2.9456723167559515e-06, + "loss": 0.0119, + "step": 96870 + }, + { + "epoch": 28.18737270875764, + "grad_norm": 0.1299230009317398, + "learning_rate": 2.9432126583664354e-06, + "loss": 0.0105, + "step": 96880 + }, + { + "epoch": 28.19028222286878, + "grad_norm": 0.08996094018220901, + "learning_rate": 2.940753915610355e-06, + "loss": 0.0127, + "step": 96890 + }, + { + "epoch": 28.193191736979923, + "grad_norm": 0.22014160454273224, + "learning_rate": 2.9382960886744264e-06, + "loss": 0.0127, + "step": 96900 + }, + { + "epoch": 28.19610125109107, + "grad_norm": 0.0880308449268341, + "learning_rate": 2.935839177745308e-06, + "loss": 0.0101, + "step": 96910 + }, + { + "epoch": 28.19901076520221, + "grad_norm": 0.12600280344486237, + "learning_rate": 2.933383183009594e-06, + "loss": 0.0117, + "step": 96920 + }, + { + "epoch": 28.201920279313356, + "grad_norm": 0.08913987129926682, + "learning_rate": 2.930928104653789e-06, + "loss": 0.0106, + "step": 96930 + }, + { + "epoch": 28.204829793424498, + "grad_norm": 0.08044512569904327, + "learning_rate": 2.92847394286435e-06, + "loss": 0.0102, + "step": 96940 + }, + { + "epoch": 28.207739307535643, + "grad_norm": 0.3798866271972656, + "learning_rate": 2.9260206978276444e-06, + "loss": 0.0141, + "step": 96950 + }, + { + "epoch": 28.210648821646785, + "grad_norm": 0.10851485282182693, + "learning_rate": 2.923568369729992e-06, + "loss": 0.0114, + "step": 96960 + }, + { + "epoch": 28.213558335757927, + "grad_norm": 0.1453690528869629, + "learning_rate": 2.9211169587576186e-06, + "loss": 0.0105, + "step": 96970 + }, + { + "epoch": 28.216467849869073, + "grad_norm": 0.08750668168067932, + "learning_rate": 2.918666465096699e-06, + "loss": 0.0115, + "step": 96980 + }, + { + "epoch": 28.219377363980215, + "grad_norm": 0.10711812227964401, + "learning_rate": 2.916216888933335e-06, + "loss": 0.0127, + "step": 96990 + }, + { + "epoch": 28.22228687809136, + "grad_norm": 0.1936221718788147, + "learning_rate": 2.9137682304535524e-06, + "loss": 0.012, + "step": 97000 + }, + { + "epoch": 28.225196392202502, + "grad_norm": 0.2614877223968506, + "learning_rate": 2.911320489843303e-06, + "loss": 0.0118, + "step": 97010 + }, + { + "epoch": 28.228105906313644, + "grad_norm": 0.13195118308067322, + "learning_rate": 2.9088736672884923e-06, + "loss": 0.0112, + "step": 97020 + }, + { + "epoch": 28.23101542042479, + "grad_norm": 0.11671222746372223, + "learning_rate": 2.9064277629749315e-06, + "loss": 0.0143, + "step": 97030 + }, + { + "epoch": 28.23392493453593, + "grad_norm": 0.1294061243534088, + "learning_rate": 2.903982777088368e-06, + "loss": 0.0102, + "step": 97040 + }, + { + "epoch": 28.236834448647077, + "grad_norm": 0.1323583424091339, + "learning_rate": 2.901538709814486e-06, + "loss": 0.0098, + "step": 97050 + }, + { + "epoch": 28.23974396275822, + "grad_norm": 0.09344936907291412, + "learning_rate": 2.8990955613388994e-06, + "loss": 0.0121, + "step": 97060 + }, + { + "epoch": 28.242653476869364, + "grad_norm": 0.11015602201223373, + "learning_rate": 2.896653331847145e-06, + "loss": 0.0114, + "step": 97070 + }, + { + "epoch": 28.245562990980506, + "grad_norm": 0.12129401415586472, + "learning_rate": 2.8942120215246916e-06, + "loss": 0.0101, + "step": 97080 + }, + { + "epoch": 28.248472505091648, + "grad_norm": 0.12382204085588455, + "learning_rate": 2.891771630556945e-06, + "loss": 0.0124, + "step": 97090 + }, + { + "epoch": 28.251382019202794, + "grad_norm": 0.09607204794883728, + "learning_rate": 2.889332159129238e-06, + "loss": 0.0089, + "step": 97100 + }, + { + "epoch": 28.254291533313936, + "grad_norm": 0.11401520669460297, + "learning_rate": 2.8868936074268322e-06, + "loss": 0.0097, + "step": 97110 + }, + { + "epoch": 28.25720104742508, + "grad_norm": 0.06479570269584656, + "learning_rate": 2.8844559756349073e-06, + "loss": 0.0105, + "step": 97120 + }, + { + "epoch": 28.260110561536223, + "grad_norm": 0.09403529763221741, + "learning_rate": 2.882019263938605e-06, + "loss": 0.0103, + "step": 97130 + }, + { + "epoch": 28.26302007564737, + "grad_norm": 0.10944997519254684, + "learning_rate": 2.8795834725229663e-06, + "loss": 0.0101, + "step": 97140 + }, + { + "epoch": 28.26592958975851, + "grad_norm": 0.11742119491100311, + "learning_rate": 2.8771486015729733e-06, + "loss": 0.0117, + "step": 97150 + }, + { + "epoch": 28.268839103869652, + "grad_norm": 0.23130394518375397, + "learning_rate": 2.874714651273537e-06, + "loss": 0.0115, + "step": 97160 + }, + { + "epoch": 28.271748617980798, + "grad_norm": 0.17706042528152466, + "learning_rate": 2.8722816218095095e-06, + "loss": 0.0127, + "step": 97170 + }, + { + "epoch": 28.27465813209194, + "grad_norm": 0.08700937777757645, + "learning_rate": 2.869849513365653e-06, + "loss": 0.0115, + "step": 97180 + }, + { + "epoch": 28.277567646203085, + "grad_norm": 0.1070399284362793, + "learning_rate": 2.8674183261266764e-06, + "loss": 0.0103, + "step": 97190 + }, + { + "epoch": 28.280477160314227, + "grad_norm": 0.10712640732526779, + "learning_rate": 2.864988060277207e-06, + "loss": 0.0105, + "step": 97200 + }, + { + "epoch": 28.283386674425373, + "grad_norm": 0.08952908217906952, + "learning_rate": 2.8625587160018136e-06, + "loss": 0.011, + "step": 97210 + }, + { + "epoch": 28.286296188536515, + "grad_norm": 0.22029472887516022, + "learning_rate": 2.8601302934849827e-06, + "loss": 0.0106, + "step": 97220 + }, + { + "epoch": 28.289205702647656, + "grad_norm": 0.18591037392616272, + "learning_rate": 2.8577027929111387e-06, + "loss": 0.0127, + "step": 97230 + }, + { + "epoch": 28.292115216758802, + "grad_norm": 0.1436394900083542, + "learning_rate": 2.8552762144646394e-06, + "loss": 0.0108, + "step": 97240 + }, + { + "epoch": 28.295024730869944, + "grad_norm": 0.09955525398254395, + "learning_rate": 2.8528505583297603e-06, + "loss": 0.0111, + "step": 97250 + }, + { + "epoch": 28.29793424498109, + "grad_norm": 0.10290397703647614, + "learning_rate": 2.8504258246907193e-06, + "loss": 0.0133, + "step": 97260 + }, + { + "epoch": 28.30084375909223, + "grad_norm": 0.12144531309604645, + "learning_rate": 2.8480020137316544e-06, + "loss": 0.011, + "step": 97270 + }, + { + "epoch": 28.303753273203377, + "grad_norm": 0.16139477491378784, + "learning_rate": 2.845579125636639e-06, + "loss": 0.0131, + "step": 97280 + }, + { + "epoch": 28.30666278731452, + "grad_norm": 0.11221969872713089, + "learning_rate": 2.843157160589682e-06, + "loss": 0.0101, + "step": 97290 + }, + { + "epoch": 28.30957230142566, + "grad_norm": 0.12737680971622467, + "learning_rate": 2.8407361187747113e-06, + "loss": 0.0121, + "step": 97300 + }, + { + "epoch": 28.312481815536806, + "grad_norm": 0.084770567715168, + "learning_rate": 2.838316000375582e-06, + "loss": 0.0101, + "step": 97310 + }, + { + "epoch": 28.315391329647948, + "grad_norm": 0.12830707430839539, + "learning_rate": 2.835896805576093e-06, + "loss": 0.0132, + "step": 97320 + }, + { + "epoch": 28.318300843759094, + "grad_norm": 0.09640057384967804, + "learning_rate": 2.8334785345599704e-06, + "loss": 0.0129, + "step": 97330 + }, + { + "epoch": 28.321210357870235, + "grad_norm": 0.07978063821792603, + "learning_rate": 2.831061187510856e-06, + "loss": 0.0098, + "step": 97340 + }, + { + "epoch": 28.324119871981377, + "grad_norm": 0.14277233183383942, + "learning_rate": 2.8286447646123376e-06, + "loss": 0.0126, + "step": 97350 + }, + { + "epoch": 28.327029386092523, + "grad_norm": 0.13635756075382233, + "learning_rate": 2.8262292660479283e-06, + "loss": 0.0134, + "step": 97360 + }, + { + "epoch": 28.329938900203665, + "grad_norm": 0.1713048815727234, + "learning_rate": 2.8238146920010667e-06, + "loss": 0.0105, + "step": 97370 + }, + { + "epoch": 28.33284841431481, + "grad_norm": 0.1520623415708542, + "learning_rate": 2.8214010426551207e-06, + "loss": 0.0125, + "step": 97380 + }, + { + "epoch": 28.335757928425952, + "grad_norm": 0.13453158736228943, + "learning_rate": 2.8189883181933922e-06, + "loss": 0.0117, + "step": 97390 + }, + { + "epoch": 28.338667442537098, + "grad_norm": 0.08837773650884628, + "learning_rate": 2.8165765187991167e-06, + "loss": 0.0102, + "step": 97400 + }, + { + "epoch": 28.34157695664824, + "grad_norm": 0.06950005143880844, + "learning_rate": 2.814165644655452e-06, + "loss": 0.0102, + "step": 97410 + }, + { + "epoch": 28.34448647075938, + "grad_norm": 0.15676435828208923, + "learning_rate": 2.8117556959454787e-06, + "loss": 0.0102, + "step": 97420 + }, + { + "epoch": 28.347395984870527, + "grad_norm": 0.11227737367153168, + "learning_rate": 2.809346672852234e-06, + "loss": 0.0103, + "step": 97430 + }, + { + "epoch": 28.35030549898167, + "grad_norm": 0.10188087075948715, + "learning_rate": 2.8069385755586572e-06, + "loss": 0.0118, + "step": 97440 + }, + { + "epoch": 28.353215013092814, + "grad_norm": 0.11687441915273666, + "learning_rate": 2.804531404247625e-06, + "loss": 0.0116, + "step": 97450 + }, + { + "epoch": 28.356124527203956, + "grad_norm": 0.08592475950717926, + "learning_rate": 2.8021251591019486e-06, + "loss": 0.0097, + "step": 97460 + }, + { + "epoch": 28.359034041315102, + "grad_norm": 0.13374298810958862, + "learning_rate": 2.7997198403043724e-06, + "loss": 0.0103, + "step": 97470 + }, + { + "epoch": 28.361943555426244, + "grad_norm": 0.09212581813335419, + "learning_rate": 2.7973154480375545e-06, + "loss": 0.0123, + "step": 97480 + }, + { + "epoch": 28.364853069537386, + "grad_norm": 0.27165988087654114, + "learning_rate": 2.794911982484101e-06, + "loss": 0.0103, + "step": 97490 + }, + { + "epoch": 28.36776258364853, + "grad_norm": 0.178179532289505, + "learning_rate": 2.792509443826533e-06, + "loss": 0.0115, + "step": 97500 + }, + { + "epoch": 28.370672097759673, + "grad_norm": 0.5642987489700317, + "learning_rate": 2.790107832247314e-06, + "loss": 0.0099, + "step": 97510 + }, + { + "epoch": 28.37358161187082, + "grad_norm": 0.1536235511302948, + "learning_rate": 2.7877071479288248e-06, + "loss": 0.0109, + "step": 97520 + }, + { + "epoch": 28.37649112598196, + "grad_norm": 0.14793570339679718, + "learning_rate": 2.785307391053381e-06, + "loss": 0.0111, + "step": 97530 + }, + { + "epoch": 28.379400640093106, + "grad_norm": 0.09213881939649582, + "learning_rate": 2.7829085618032355e-06, + "loss": 0.0115, + "step": 97540 + }, + { + "epoch": 28.382310154204248, + "grad_norm": 0.08098451048135757, + "learning_rate": 2.780510660360556e-06, + "loss": 0.0109, + "step": 97550 + }, + { + "epoch": 28.38521966831539, + "grad_norm": 0.14281107485294342, + "learning_rate": 2.7781136869074535e-06, + "loss": 0.0113, + "step": 97560 + }, + { + "epoch": 28.388129182426535, + "grad_norm": 0.11939889937639236, + "learning_rate": 2.775717641625954e-06, + "loss": 0.0116, + "step": 97570 + }, + { + "epoch": 28.391038696537677, + "grad_norm": 0.08694462478160858, + "learning_rate": 2.7733225246980315e-06, + "loss": 0.0101, + "step": 97580 + }, + { + "epoch": 28.393948210648823, + "grad_norm": 0.08746728301048279, + "learning_rate": 2.770928336305571e-06, + "loss": 0.0119, + "step": 97590 + }, + { + "epoch": 28.396857724759965, + "grad_norm": 0.11682799458503723, + "learning_rate": 2.7685350766304e-06, + "loss": 0.0107, + "step": 97600 + }, + { + "epoch": 28.39976723887111, + "grad_norm": 0.13480263948440552, + "learning_rate": 2.766142745854267e-06, + "loss": 0.0109, + "step": 97610 + }, + { + "epoch": 28.402676752982252, + "grad_norm": 0.09539345651865005, + "learning_rate": 2.763751344158856e-06, + "loss": 0.0093, + "step": 97620 + }, + { + "epoch": 28.405586267093394, + "grad_norm": 0.09372766315937042, + "learning_rate": 2.7613608717257827e-06, + "loss": 0.0095, + "step": 97630 + }, + { + "epoch": 28.40849578120454, + "grad_norm": 0.06015473231673241, + "learning_rate": 2.7589713287365794e-06, + "loss": 0.0095, + "step": 97640 + }, + { + "epoch": 28.41140529531568, + "grad_norm": 0.07191997766494751, + "learning_rate": 2.756582715372719e-06, + "loss": 0.0129, + "step": 97650 + }, + { + "epoch": 28.414314809426827, + "grad_norm": 0.1006719097495079, + "learning_rate": 2.7541950318156056e-06, + "loss": 0.0111, + "step": 97660 + }, + { + "epoch": 28.41722432353797, + "grad_norm": 0.09715109318494797, + "learning_rate": 2.7518082782465635e-06, + "loss": 0.012, + "step": 97670 + }, + { + "epoch": 28.42013383764911, + "grad_norm": 0.12458264827728271, + "learning_rate": 2.74942245484685e-06, + "loss": 0.0096, + "step": 97680 + }, + { + "epoch": 28.423043351760256, + "grad_norm": 0.11561909317970276, + "learning_rate": 2.747037561797653e-06, + "loss": 0.0099, + "step": 97690 + }, + { + "epoch": 28.4259528658714, + "grad_norm": 0.095987968146801, + "learning_rate": 2.744653599280094e-06, + "loss": 0.0098, + "step": 97700 + }, + { + "epoch": 28.428862379982544, + "grad_norm": 0.1920228898525238, + "learning_rate": 2.7422705674752156e-06, + "loss": 0.0126, + "step": 97710 + }, + { + "epoch": 28.431771894093686, + "grad_norm": 0.09745332598686218, + "learning_rate": 2.7398884665639877e-06, + "loss": 0.0106, + "step": 97720 + }, + { + "epoch": 28.43468140820483, + "grad_norm": 0.10342087596654892, + "learning_rate": 2.7375072967273275e-06, + "loss": 0.0117, + "step": 97730 + }, + { + "epoch": 28.437590922315973, + "grad_norm": 0.10566645860671997, + "learning_rate": 2.735127058146063e-06, + "loss": 0.0097, + "step": 97740 + }, + { + "epoch": 28.440500436427115, + "grad_norm": 0.09446866810321808, + "learning_rate": 2.7327477510009536e-06, + "loss": 0.0111, + "step": 97750 + }, + { + "epoch": 28.44340995053826, + "grad_norm": 0.10791753232479095, + "learning_rate": 2.730369375472696e-06, + "loss": 0.0105, + "step": 97760 + }, + { + "epoch": 28.446319464649402, + "grad_norm": 0.08652960509061813, + "learning_rate": 2.727991931741916e-06, + "loss": 0.009, + "step": 97770 + }, + { + "epoch": 28.449228978760548, + "grad_norm": 0.12296925485134125, + "learning_rate": 2.725615419989159e-06, + "loss": 0.0126, + "step": 97780 + }, + { + "epoch": 28.45213849287169, + "grad_norm": 0.17461329698562622, + "learning_rate": 2.7232398403949043e-06, + "loss": 0.014, + "step": 97790 + }, + { + "epoch": 28.455048006982835, + "grad_norm": 0.10400432348251343, + "learning_rate": 2.720865193139564e-06, + "loss": 0.0091, + "step": 97800 + }, + { + "epoch": 28.457957521093977, + "grad_norm": 0.10892391949892044, + "learning_rate": 2.7184914784034804e-06, + "loss": 0.0094, + "step": 97810 + }, + { + "epoch": 28.46086703520512, + "grad_norm": 0.08354543894529343, + "learning_rate": 2.7161186963669145e-06, + "loss": 0.0121, + "step": 97820 + }, + { + "epoch": 28.463776549316265, + "grad_norm": 0.10773086547851562, + "learning_rate": 2.713746847210067e-06, + "loss": 0.0128, + "step": 97830 + }, + { + "epoch": 28.466686063427407, + "grad_norm": 0.14034081995487213, + "learning_rate": 2.7113759311130688e-06, + "loss": 0.0112, + "step": 97840 + }, + { + "epoch": 28.469595577538552, + "grad_norm": 0.07994251698255539, + "learning_rate": 2.709005948255966e-06, + "loss": 0.0085, + "step": 97850 + }, + { + "epoch": 28.472505091649694, + "grad_norm": 0.15023356676101685, + "learning_rate": 2.706636898818753e-06, + "loss": 0.0089, + "step": 97860 + }, + { + "epoch": 28.47541460576084, + "grad_norm": 0.10449380427598953, + "learning_rate": 2.7042687829813323e-06, + "loss": 0.0119, + "step": 97870 + }, + { + "epoch": 28.47832411987198, + "grad_norm": 0.130879208445549, + "learning_rate": 2.7019016009235588e-06, + "loss": 0.0106, + "step": 97880 + }, + { + "epoch": 28.481233633983123, + "grad_norm": 0.12427414953708649, + "learning_rate": 2.699535352825193e-06, + "loss": 0.0104, + "step": 97890 + }, + { + "epoch": 28.48414314809427, + "grad_norm": 0.11723575741052628, + "learning_rate": 2.6971700388659452e-06, + "loss": 0.0122, + "step": 97900 + }, + { + "epoch": 28.48705266220541, + "grad_norm": 0.10459046065807343, + "learning_rate": 2.694805659225438e-06, + "loss": 0.0112, + "step": 97910 + }, + { + "epoch": 28.489962176316556, + "grad_norm": 0.08155040442943573, + "learning_rate": 2.692442214083235e-06, + "loss": 0.0095, + "step": 97920 + }, + { + "epoch": 28.492871690427698, + "grad_norm": 0.08515691012144089, + "learning_rate": 2.690079703618825e-06, + "loss": 0.0113, + "step": 97930 + }, + { + "epoch": 28.495781204538844, + "grad_norm": 0.11430877447128296, + "learning_rate": 2.6877181280116203e-06, + "loss": 0.0092, + "step": 97940 + }, + { + "epoch": 28.498690718649986, + "grad_norm": 0.07788780331611633, + "learning_rate": 2.6853574874409736e-06, + "loss": 0.0087, + "step": 97950 + }, + { + "epoch": 28.501600232761128, + "grad_norm": 0.14842405915260315, + "learning_rate": 2.682997782086152e-06, + "loss": 0.012, + "step": 97960 + }, + { + "epoch": 28.504509746872273, + "grad_norm": 0.09185489267110825, + "learning_rate": 2.680639012126366e-06, + "loss": 0.0118, + "step": 97970 + }, + { + "epoch": 28.507419260983415, + "grad_norm": 0.22887377440929413, + "learning_rate": 2.678281177740744e-06, + "loss": 0.0121, + "step": 97980 + }, + { + "epoch": 28.51032877509456, + "grad_norm": 0.10567469894886017, + "learning_rate": 2.6759242791083518e-06, + "loss": 0.0103, + "step": 97990 + }, + { + "epoch": 28.513238289205702, + "grad_norm": 0.11717712134122849, + "learning_rate": 2.673568316408181e-06, + "loss": 0.0105, + "step": 98000 + }, + { + "epoch": 28.516147803316848, + "grad_norm": 0.10904695838689804, + "learning_rate": 2.671213289819149e-06, + "loss": 0.0119, + "step": 98010 + }, + { + "epoch": 28.51905731742799, + "grad_norm": 0.29598185420036316, + "learning_rate": 2.668859199520097e-06, + "loss": 0.0124, + "step": 98020 + }, + { + "epoch": 28.52196683153913, + "grad_norm": 0.09971896559000015, + "learning_rate": 2.666506045689818e-06, + "loss": 0.0128, + "step": 98030 + }, + { + "epoch": 28.524876345650277, + "grad_norm": 0.09877574443817139, + "learning_rate": 2.6641538285070107e-06, + "loss": 0.0114, + "step": 98040 + }, + { + "epoch": 28.52778585976142, + "grad_norm": 0.1170063465833664, + "learning_rate": 2.6618025481503077e-06, + "loss": 0.0109, + "step": 98050 + }, + { + "epoch": 28.530695373872565, + "grad_norm": 0.09255421161651611, + "learning_rate": 2.6594522047982756e-06, + "loss": 0.0103, + "step": 98060 + }, + { + "epoch": 28.533604887983707, + "grad_norm": 0.08552636951208115, + "learning_rate": 2.657102798629413e-06, + "loss": 0.0107, + "step": 98070 + }, + { + "epoch": 28.53651440209485, + "grad_norm": 0.12302565574645996, + "learning_rate": 2.654754329822135e-06, + "loss": 0.0092, + "step": 98080 + }, + { + "epoch": 28.539423916205994, + "grad_norm": 0.09225117415189743, + "learning_rate": 2.652406798554791e-06, + "loss": 0.0104, + "step": 98090 + }, + { + "epoch": 28.542333430317136, + "grad_norm": 0.10911740362644196, + "learning_rate": 2.650060205005664e-06, + "loss": 0.0096, + "step": 98100 + }, + { + "epoch": 28.54524294442828, + "grad_norm": 0.0957140177488327, + "learning_rate": 2.6477145493529652e-06, + "loss": 0.012, + "step": 98110 + }, + { + "epoch": 28.548152458539423, + "grad_norm": 0.09034126996994019, + "learning_rate": 2.6453698317748247e-06, + "loss": 0.0097, + "step": 98120 + }, + { + "epoch": 28.55106197265057, + "grad_norm": 0.10008101910352707, + "learning_rate": 2.6430260524493116e-06, + "loss": 0.0107, + "step": 98130 + }, + { + "epoch": 28.55397148676171, + "grad_norm": 0.07760139554738998, + "learning_rate": 2.6406832115544247e-06, + "loss": 0.0101, + "step": 98140 + }, + { + "epoch": 28.556881000872853, + "grad_norm": 0.08735030144453049, + "learning_rate": 2.6383413092680843e-06, + "loss": 0.0084, + "step": 98150 + }, + { + "epoch": 28.559790514983998, + "grad_norm": 0.1555217206478119, + "learning_rate": 2.6360003457681387e-06, + "loss": 0.0103, + "step": 98160 + }, + { + "epoch": 28.56270002909514, + "grad_norm": 0.083661749958992, + "learning_rate": 2.6336603212323693e-06, + "loss": 0.0104, + "step": 98170 + }, + { + "epoch": 28.565609543206286, + "grad_norm": 0.12828482687473297, + "learning_rate": 2.631321235838493e-06, + "loss": 0.0127, + "step": 98180 + }, + { + "epoch": 28.568519057317427, + "grad_norm": 0.09268027544021606, + "learning_rate": 2.6289830897641386e-06, + "loss": 0.0122, + "step": 98190 + }, + { + "epoch": 28.571428571428573, + "grad_norm": 0.15208229422569275, + "learning_rate": 2.6266458831868797e-06, + "loss": 0.0099, + "step": 98200 + }, + { + "epoch": 28.574338085539715, + "grad_norm": 0.11315120756626129, + "learning_rate": 2.624309616284207e-06, + "loss": 0.0111, + "step": 98210 + }, + { + "epoch": 28.577247599650857, + "grad_norm": 0.09908509254455566, + "learning_rate": 2.621974289233545e-06, + "loss": 0.0097, + "step": 98220 + }, + { + "epoch": 28.580157113762002, + "grad_norm": 0.1012888178229332, + "learning_rate": 2.619639902212251e-06, + "loss": 0.009, + "step": 98230 + }, + { + "epoch": 28.583066627873144, + "grad_norm": 0.09000766277313232, + "learning_rate": 2.6173064553976016e-06, + "loss": 0.0116, + "step": 98240 + }, + { + "epoch": 28.58597614198429, + "grad_norm": 0.10230209678411484, + "learning_rate": 2.6149739489668113e-06, + "loss": 0.0106, + "step": 98250 + }, + { + "epoch": 28.58888565609543, + "grad_norm": 0.0878186821937561, + "learning_rate": 2.6126423830970114e-06, + "loss": 0.0115, + "step": 98260 + }, + { + "epoch": 28.591795170206577, + "grad_norm": 0.08747760206460953, + "learning_rate": 2.6103117579652758e-06, + "loss": 0.0103, + "step": 98270 + }, + { + "epoch": 28.59470468431772, + "grad_norm": 0.10793249309062958, + "learning_rate": 2.607982073748596e-06, + "loss": 0.0107, + "step": 98280 + }, + { + "epoch": 28.59761419842886, + "grad_norm": 0.12229033559560776, + "learning_rate": 2.6056533306238955e-06, + "loss": 0.0104, + "step": 98290 + }, + { + "epoch": 28.600523712540006, + "grad_norm": 0.1567593216896057, + "learning_rate": 2.603325528768033e-06, + "loss": 0.0135, + "step": 98300 + }, + { + "epoch": 28.60343322665115, + "grad_norm": 0.12385322898626328, + "learning_rate": 2.6009986683577867e-06, + "loss": 0.0102, + "step": 98310 + }, + { + "epoch": 28.606342740762294, + "grad_norm": 0.10537239909172058, + "learning_rate": 2.598672749569861e-06, + "loss": 0.0093, + "step": 98320 + }, + { + "epoch": 28.609252254873436, + "grad_norm": 0.09883871674537659, + "learning_rate": 2.596347772580897e-06, + "loss": 0.0125, + "step": 98330 + }, + { + "epoch": 28.612161768984578, + "grad_norm": 0.11696377396583557, + "learning_rate": 2.5940237375674673e-06, + "loss": 0.0105, + "step": 98340 + }, + { + "epoch": 28.615071283095723, + "grad_norm": 0.11016026884317398, + "learning_rate": 2.5917006447060595e-06, + "loss": 0.013, + "step": 98350 + }, + { + "epoch": 28.617980797206865, + "grad_norm": 0.12948860228061676, + "learning_rate": 2.5893784941730995e-06, + "loss": 0.012, + "step": 98360 + }, + { + "epoch": 28.62089031131801, + "grad_norm": 0.05309295281767845, + "learning_rate": 2.5870572861449443e-06, + "loss": 0.0101, + "step": 98370 + }, + { + "epoch": 28.623799825429153, + "grad_norm": 0.11482275277376175, + "learning_rate": 2.5847370207978687e-06, + "loss": 0.009, + "step": 98380 + }, + { + "epoch": 28.626709339540298, + "grad_norm": 0.09720452129840851, + "learning_rate": 2.5824176983080804e-06, + "loss": 0.0105, + "step": 98390 + }, + { + "epoch": 28.62961885365144, + "grad_norm": 0.08250757306814194, + "learning_rate": 2.580099318851717e-06, + "loss": 0.0109, + "step": 98400 + }, + { + "epoch": 28.632528367762582, + "grad_norm": 0.113948754966259, + "learning_rate": 2.577781882604852e-06, + "loss": 0.0128, + "step": 98410 + }, + { + "epoch": 28.635437881873727, + "grad_norm": 0.1073017418384552, + "learning_rate": 2.575465389743472e-06, + "loss": 0.0119, + "step": 98420 + }, + { + "epoch": 28.63834739598487, + "grad_norm": 0.09532332420349121, + "learning_rate": 2.5731498404434945e-06, + "loss": 0.0106, + "step": 98430 + }, + { + "epoch": 28.641256910096015, + "grad_norm": 0.23106525838375092, + "learning_rate": 2.5708352348807835e-06, + "loss": 0.0106, + "step": 98440 + }, + { + "epoch": 28.644166424207157, + "grad_norm": 0.1134149432182312, + "learning_rate": 2.5685215732311103e-06, + "loss": 0.0095, + "step": 98450 + }, + { + "epoch": 28.647075938318302, + "grad_norm": 0.08352121710777283, + "learning_rate": 2.5662088556701807e-06, + "loss": 0.0091, + "step": 98460 + }, + { + "epoch": 28.649985452429444, + "grad_norm": 0.10607556998729706, + "learning_rate": 2.5638970823736312e-06, + "loss": 0.011, + "step": 98470 + }, + { + "epoch": 28.652894966540586, + "grad_norm": 0.08597410470247269, + "learning_rate": 2.5615862535170316e-06, + "loss": 0.0108, + "step": 98480 + }, + { + "epoch": 28.65580448065173, + "grad_norm": 0.10640964657068253, + "learning_rate": 2.559276369275866e-06, + "loss": 0.0117, + "step": 98490 + }, + { + "epoch": 28.658713994762874, + "grad_norm": 0.06963127851486206, + "learning_rate": 2.5569674298255625e-06, + "loss": 0.0112, + "step": 98500 + }, + { + "epoch": 28.66162350887402, + "grad_norm": 0.09400637447834015, + "learning_rate": 2.5546594353414625e-06, + "loss": 0.0108, + "step": 98510 + }, + { + "epoch": 28.66453302298516, + "grad_norm": 0.10921163856983185, + "learning_rate": 2.552352385998848e-06, + "loss": 0.0112, + "step": 98520 + }, + { + "epoch": 28.667442537096306, + "grad_norm": 0.12310309708118439, + "learning_rate": 2.5500462819729213e-06, + "loss": 0.0112, + "step": 98530 + }, + { + "epoch": 28.67035205120745, + "grad_norm": 0.1397581696510315, + "learning_rate": 2.5477411234388147e-06, + "loss": 0.012, + "step": 98540 + }, + { + "epoch": 28.67326156531859, + "grad_norm": 0.09340006858110428, + "learning_rate": 2.5454369105715965e-06, + "loss": 0.0112, + "step": 98550 + }, + { + "epoch": 28.676171079429736, + "grad_norm": 0.08621671050786972, + "learning_rate": 2.5431336435462494e-06, + "loss": 0.0107, + "step": 98560 + }, + { + "epoch": 28.679080593540878, + "grad_norm": 0.12417156994342804, + "learning_rate": 2.5408313225376955e-06, + "loss": 0.0111, + "step": 98570 + }, + { + "epoch": 28.681990107652023, + "grad_norm": 0.0848313719034195, + "learning_rate": 2.538529947720777e-06, + "loss": 0.0115, + "step": 98580 + }, + { + "epoch": 28.684899621763165, + "grad_norm": 0.1921319216489792, + "learning_rate": 2.53622951927027e-06, + "loss": 0.0118, + "step": 98590 + }, + { + "epoch": 28.68780913587431, + "grad_norm": 0.07869099825620651, + "learning_rate": 2.533930037360882e-06, + "loss": 0.0105, + "step": 98600 + }, + { + "epoch": 28.690718649985453, + "grad_norm": 0.09091523289680481, + "learning_rate": 2.5316315021672384e-06, + "loss": 0.0108, + "step": 98610 + }, + { + "epoch": 28.693628164096594, + "grad_norm": 0.09435641765594482, + "learning_rate": 2.5293339138638937e-06, + "loss": 0.0103, + "step": 98620 + }, + { + "epoch": 28.69653767820774, + "grad_norm": 0.10650788247585297, + "learning_rate": 2.5270372726253392e-06, + "loss": 0.0106, + "step": 98630 + }, + { + "epoch": 28.699447192318882, + "grad_norm": 0.08433486521244049, + "learning_rate": 2.5247415786259938e-06, + "loss": 0.0108, + "step": 98640 + }, + { + "epoch": 28.702356706430027, + "grad_norm": 0.11160383373498917, + "learning_rate": 2.5224468320401907e-06, + "loss": 0.0104, + "step": 98650 + }, + { + "epoch": 28.70526622054117, + "grad_norm": 0.07961607724428177, + "learning_rate": 2.5201530330422075e-06, + "loss": 0.0124, + "step": 98660 + }, + { + "epoch": 28.708175734652315, + "grad_norm": 0.12794265151023865, + "learning_rate": 2.517860181806245e-06, + "loss": 0.0112, + "step": 98670 + }, + { + "epoch": 28.711085248763457, + "grad_norm": 0.13509564101696014, + "learning_rate": 2.5155682785064266e-06, + "loss": 0.0118, + "step": 98680 + }, + { + "epoch": 28.7139947628746, + "grad_norm": 0.0991855338215828, + "learning_rate": 2.5132773233168006e-06, + "loss": 0.0109, + "step": 98690 + }, + { + "epoch": 28.716904276985744, + "grad_norm": 0.10036944597959518, + "learning_rate": 2.5109873164113623e-06, + "loss": 0.0126, + "step": 98700 + }, + { + "epoch": 28.719813791096886, + "grad_norm": 0.14113402366638184, + "learning_rate": 2.5086982579640177e-06, + "loss": 0.0109, + "step": 98710 + }, + { + "epoch": 28.72272330520803, + "grad_norm": 0.09047998487949371, + "learning_rate": 2.5064101481486062e-06, + "loss": 0.0096, + "step": 98720 + }, + { + "epoch": 28.725632819319173, + "grad_norm": 0.12691077589988708, + "learning_rate": 2.5041229871388854e-06, + "loss": 0.0135, + "step": 98730 + }, + { + "epoch": 28.72854233343032, + "grad_norm": 0.09234873205423355, + "learning_rate": 2.5018367751085647e-06, + "loss": 0.0098, + "step": 98740 + }, + { + "epoch": 28.73145184754146, + "grad_norm": 0.11564215272665024, + "learning_rate": 2.4995515122312616e-06, + "loss": 0.0103, + "step": 98750 + }, + { + "epoch": 28.734361361652603, + "grad_norm": 0.09374123066663742, + "learning_rate": 2.497267198680521e-06, + "loss": 0.0113, + "step": 98760 + }, + { + "epoch": 28.73727087576375, + "grad_norm": 0.1096271201968193, + "learning_rate": 2.494983834629826e-06, + "loss": 0.0097, + "step": 98770 + }, + { + "epoch": 28.74018038987489, + "grad_norm": 0.13074736297130585, + "learning_rate": 2.4927014202525856e-06, + "loss": 0.0098, + "step": 98780 + }, + { + "epoch": 28.743089903986036, + "grad_norm": 0.09975916147232056, + "learning_rate": 2.490419955722133e-06, + "loss": 0.0115, + "step": 98790 + }, + { + "epoch": 28.745999418097178, + "grad_norm": 0.16687904298305511, + "learning_rate": 2.48813944121172e-06, + "loss": 0.0108, + "step": 98800 + }, + { + "epoch": 28.74890893220832, + "grad_norm": 0.1188945472240448, + "learning_rate": 2.4858598768945523e-06, + "loss": 0.0101, + "step": 98810 + }, + { + "epoch": 28.751818446319465, + "grad_norm": 0.10018423944711685, + "learning_rate": 2.483581262943742e-06, + "loss": 0.011, + "step": 98820 + }, + { + "epoch": 28.754727960430607, + "grad_norm": 0.11326644569635391, + "learning_rate": 2.481303599532329e-06, + "loss": 0.0136, + "step": 98830 + }, + { + "epoch": 28.757637474541752, + "grad_norm": 0.1016308069229126, + "learning_rate": 2.4790268868332935e-06, + "loss": 0.0102, + "step": 98840 + }, + { + "epoch": 28.760546988652894, + "grad_norm": 0.09539709240198135, + "learning_rate": 2.476751125019536e-06, + "loss": 0.0102, + "step": 98850 + }, + { + "epoch": 28.76345650276404, + "grad_norm": 0.12865202128887177, + "learning_rate": 2.4744763142638817e-06, + "loss": 0.01, + "step": 98860 + }, + { + "epoch": 28.766366016875182, + "grad_norm": 0.11082407832145691, + "learning_rate": 2.4722024547390926e-06, + "loss": 0.0119, + "step": 98870 + }, + { + "epoch": 28.769275530986324, + "grad_norm": 0.10277251899242401, + "learning_rate": 2.4699295466178496e-06, + "loss": 0.0101, + "step": 98880 + }, + { + "epoch": 28.77218504509747, + "grad_norm": 0.11792582273483276, + "learning_rate": 2.4676575900727675e-06, + "loss": 0.0109, + "step": 98890 + }, + { + "epoch": 28.77509455920861, + "grad_norm": 0.08815564960241318, + "learning_rate": 2.4653865852763823e-06, + "loss": 0.0122, + "step": 98900 + }, + { + "epoch": 28.778004073319757, + "grad_norm": 0.09326225519180298, + "learning_rate": 2.463116532401166e-06, + "loss": 0.0116, + "step": 98910 + }, + { + "epoch": 28.7809135874309, + "grad_norm": 0.14395244419574738, + "learning_rate": 2.4608474316195144e-06, + "loss": 0.0106, + "step": 98920 + }, + { + "epoch": 28.783823101542044, + "grad_norm": 0.0867842584848404, + "learning_rate": 2.4585792831037463e-06, + "loss": 0.0108, + "step": 98930 + }, + { + "epoch": 28.786732615653186, + "grad_norm": 0.15091918408870697, + "learning_rate": 2.4563120870261195e-06, + "loss": 0.0127, + "step": 98940 + }, + { + "epoch": 28.789642129764328, + "grad_norm": 0.08363914489746094, + "learning_rate": 2.454045843558804e-06, + "loss": 0.0109, + "step": 98950 + }, + { + "epoch": 28.792551643875473, + "grad_norm": 0.11258859187364578, + "learning_rate": 2.4517805528739107e-06, + "loss": 0.0105, + "step": 98960 + }, + { + "epoch": 28.795461157986615, + "grad_norm": 0.14858020842075348, + "learning_rate": 2.449516215143478e-06, + "loss": 0.0102, + "step": 98970 + }, + { + "epoch": 28.79837067209776, + "grad_norm": 0.06824581325054169, + "learning_rate": 2.447252830539463e-06, + "loss": 0.0101, + "step": 98980 + }, + { + "epoch": 28.801280186208903, + "grad_norm": 0.10020966082811356, + "learning_rate": 2.44499039923375e-06, + "loss": 0.0088, + "step": 98990 + }, + { + "epoch": 28.804189700320048, + "grad_norm": 0.10217002034187317, + "learning_rate": 2.4427289213981602e-06, + "loss": 0.0117, + "step": 99000 + }, + { + "epoch": 28.80709921443119, + "grad_norm": 0.07805261760950089, + "learning_rate": 2.440468397204442e-06, + "loss": 0.012, + "step": 99010 + }, + { + "epoch": 28.810008728542332, + "grad_norm": 0.07987900823354721, + "learning_rate": 2.438208826824259e-06, + "loss": 0.0114, + "step": 99020 + }, + { + "epoch": 28.812918242653478, + "grad_norm": 0.11394568532705307, + "learning_rate": 2.435950210429217e-06, + "loss": 0.0108, + "step": 99030 + }, + { + "epoch": 28.81582775676462, + "grad_norm": 0.11924725025892258, + "learning_rate": 2.4336925481908435e-06, + "loss": 0.0111, + "step": 99040 + }, + { + "epoch": 28.818737270875765, + "grad_norm": 0.09126931428909302, + "learning_rate": 2.43143584028059e-06, + "loss": 0.0093, + "step": 99050 + }, + { + "epoch": 28.821646784986907, + "grad_norm": 0.07232232391834259, + "learning_rate": 2.4291800868698354e-06, + "loss": 0.0115, + "step": 99060 + }, + { + "epoch": 28.82455629909805, + "grad_norm": 0.11351070553064346, + "learning_rate": 2.4269252881298936e-06, + "loss": 0.0106, + "step": 99070 + }, + { + "epoch": 28.827465813209194, + "grad_norm": 0.15333552658557892, + "learning_rate": 2.4246714442320057e-06, + "loss": 0.0107, + "step": 99080 + }, + { + "epoch": 28.830375327320336, + "grad_norm": 0.16162265837192535, + "learning_rate": 2.422418555347329e-06, + "loss": 0.0085, + "step": 99090 + }, + { + "epoch": 28.83328484143148, + "grad_norm": 0.10288070887327194, + "learning_rate": 2.4201666216469533e-06, + "loss": 0.011, + "step": 99100 + }, + { + "epoch": 28.836194355542624, + "grad_norm": 0.1820330023765564, + "learning_rate": 2.41791564330191e-06, + "loss": 0.0133, + "step": 99110 + }, + { + "epoch": 28.83910386965377, + "grad_norm": 0.09438538551330566, + "learning_rate": 2.4156656204831385e-06, + "loss": 0.0105, + "step": 99120 + }, + { + "epoch": 28.84201338376491, + "grad_norm": 0.0913604274392128, + "learning_rate": 2.4134165533615097e-06, + "loss": 0.0102, + "step": 99130 + }, + { + "epoch": 28.844922897876053, + "grad_norm": 0.1055363342165947, + "learning_rate": 2.4111684421078294e-06, + "loss": 0.0109, + "step": 99140 + }, + { + "epoch": 28.8478324119872, + "grad_norm": 0.11891163885593414, + "learning_rate": 2.4089212868928323e-06, + "loss": 0.0119, + "step": 99150 + }, + { + "epoch": 28.85074192609834, + "grad_norm": 0.0696154460310936, + "learning_rate": 2.4066750878871668e-06, + "loss": 0.0112, + "step": 99160 + }, + { + "epoch": 28.853651440209486, + "grad_norm": 0.0779939815402031, + "learning_rate": 2.4044298452614185e-06, + "loss": 0.0109, + "step": 99170 + }, + { + "epoch": 28.856560954320628, + "grad_norm": 0.08058398962020874, + "learning_rate": 2.4021855591860985e-06, + "loss": 0.0093, + "step": 99180 + }, + { + "epoch": 28.859470468431773, + "grad_norm": 0.1437763273715973, + "learning_rate": 2.3999422298316515e-06, + "loss": 0.0101, + "step": 99190 + }, + { + "epoch": 28.862379982542915, + "grad_norm": 0.1369076669216156, + "learning_rate": 2.3976998573684353e-06, + "loss": 0.0122, + "step": 99200 + }, + { + "epoch": 28.865289496654057, + "grad_norm": 0.12028560042381287, + "learning_rate": 2.3954584419667463e-06, + "loss": 0.0095, + "step": 99210 + }, + { + "epoch": 28.868199010765203, + "grad_norm": 0.10009239614009857, + "learning_rate": 2.3932179837968093e-06, + "loss": 0.0124, + "step": 99220 + }, + { + "epoch": 28.871108524876345, + "grad_norm": 0.16398055851459503, + "learning_rate": 2.3909784830287666e-06, + "loss": 0.013, + "step": 99230 + }, + { + "epoch": 28.87401803898749, + "grad_norm": 0.10879532247781754, + "learning_rate": 2.3887399398326976e-06, + "loss": 0.0126, + "step": 99240 + }, + { + "epoch": 28.876927553098632, + "grad_norm": 0.17434367537498474, + "learning_rate": 2.3865023543786007e-06, + "loss": 0.0119, + "step": 99250 + }, + { + "epoch": 28.879837067209778, + "grad_norm": 0.12708435952663422, + "learning_rate": 2.384265726836412e-06, + "loss": 0.0117, + "step": 99260 + }, + { + "epoch": 28.88274658132092, + "grad_norm": 0.09088437259197235, + "learning_rate": 2.382030057375981e-06, + "loss": 0.0096, + "step": 99270 + }, + { + "epoch": 28.88565609543206, + "grad_norm": 0.08715971559286118, + "learning_rate": 2.379795346167099e-06, + "loss": 0.0105, + "step": 99280 + }, + { + "epoch": 28.888565609543207, + "grad_norm": 0.08041825890541077, + "learning_rate": 2.3775615933794697e-06, + "loss": 0.01, + "step": 99290 + }, + { + "epoch": 28.89147512365435, + "grad_norm": 0.1375250518321991, + "learning_rate": 2.375328799182739e-06, + "loss": 0.0137, + "step": 99300 + }, + { + "epoch": 28.894384637765494, + "grad_norm": 0.12619927525520325, + "learning_rate": 2.3730969637464722e-06, + "loss": 0.0125, + "step": 99310 + }, + { + "epoch": 28.897294151876636, + "grad_norm": 0.11506721377372742, + "learning_rate": 2.370866087240158e-06, + "loss": 0.0105, + "step": 99320 + }, + { + "epoch": 28.90020366598778, + "grad_norm": 0.10955164581537247, + "learning_rate": 2.368636169833218e-06, + "loss": 0.0102, + "step": 99330 + }, + { + "epoch": 28.903113180098924, + "grad_norm": 0.08894036710262299, + "learning_rate": 2.3664072116950065e-06, + "loss": 0.01, + "step": 99340 + }, + { + "epoch": 28.906022694210066, + "grad_norm": 0.12434007227420807, + "learning_rate": 2.364179212994792e-06, + "loss": 0.0114, + "step": 99350 + }, + { + "epoch": 28.90893220832121, + "grad_norm": 0.10362933576107025, + "learning_rate": 2.361952173901773e-06, + "loss": 0.0114, + "step": 99360 + }, + { + "epoch": 28.911841722432353, + "grad_norm": 0.10673109441995621, + "learning_rate": 2.3597260945850842e-06, + "loss": 0.01, + "step": 99370 + }, + { + "epoch": 28.9147512365435, + "grad_norm": 0.12102537602186203, + "learning_rate": 2.3575009752137815e-06, + "loss": 0.0115, + "step": 99380 + }, + { + "epoch": 28.91766075065464, + "grad_norm": 0.35487326979637146, + "learning_rate": 2.3552768159568486e-06, + "loss": 0.0122, + "step": 99390 + }, + { + "epoch": 28.920570264765786, + "grad_norm": 0.09448620676994324, + "learning_rate": 2.3530536169831857e-06, + "loss": 0.0085, + "step": 99400 + }, + { + "epoch": 28.923479778876928, + "grad_norm": 0.12771224975585938, + "learning_rate": 2.3508313784616437e-06, + "loss": 0.0115, + "step": 99410 + }, + { + "epoch": 28.92638929298807, + "grad_norm": 0.095389224588871, + "learning_rate": 2.3486101005609828e-06, + "loss": 0.0126, + "step": 99420 + }, + { + "epoch": 28.929298807099215, + "grad_norm": 0.0789559856057167, + "learning_rate": 2.346389783449888e-06, + "loss": 0.0101, + "step": 99430 + }, + { + "epoch": 28.932208321210357, + "grad_norm": 0.09321089833974838, + "learning_rate": 2.344170427296985e-06, + "loss": 0.0103, + "step": 99440 + }, + { + "epoch": 28.935117835321503, + "grad_norm": 0.129488006234169, + "learning_rate": 2.3419520322708183e-06, + "loss": 0.0128, + "step": 99450 + }, + { + "epoch": 28.938027349432645, + "grad_norm": 0.07236841320991516, + "learning_rate": 2.33973459853986e-06, + "loss": 0.0121, + "step": 99460 + }, + { + "epoch": 28.940936863543786, + "grad_norm": 0.10917532444000244, + "learning_rate": 2.3375181262725036e-06, + "loss": 0.0122, + "step": 99470 + }, + { + "epoch": 28.943846377654932, + "grad_norm": 0.10473567247390747, + "learning_rate": 2.3353026156370816e-06, + "loss": 0.0093, + "step": 99480 + }, + { + "epoch": 28.946755891766074, + "grad_norm": 0.10416476428508759, + "learning_rate": 2.333088066801849e-06, + "loss": 0.0128, + "step": 99490 + }, + { + "epoch": 28.94966540587722, + "grad_norm": 0.1001383364200592, + "learning_rate": 2.33087447993498e-06, + "loss": 0.0118, + "step": 99500 + }, + { + "epoch": 28.95257491998836, + "grad_norm": 0.08039190620183945, + "learning_rate": 2.3286618552045856e-06, + "loss": 0.0092, + "step": 99510 + }, + { + "epoch": 28.955484434099507, + "grad_norm": 0.13723987340927124, + "learning_rate": 2.326450192778703e-06, + "loss": 0.0122, + "step": 99520 + }, + { + "epoch": 28.95839394821065, + "grad_norm": 0.11497245728969574, + "learning_rate": 2.3242394928252903e-06, + "loss": 0.0122, + "step": 99530 + }, + { + "epoch": 28.96130346232179, + "grad_norm": 0.15377052128314972, + "learning_rate": 2.3220297555122317e-06, + "loss": 0.0124, + "step": 99540 + }, + { + "epoch": 28.964212976432936, + "grad_norm": 0.14190872013568878, + "learning_rate": 2.3198209810073463e-06, + "loss": 0.0114, + "step": 99550 + }, + { + "epoch": 28.967122490544078, + "grad_norm": 0.101814866065979, + "learning_rate": 2.3176131694783782e-06, + "loss": 0.0104, + "step": 99560 + }, + { + "epoch": 28.970032004655224, + "grad_norm": 0.10409971326589584, + "learning_rate": 2.31540632109299e-06, + "loss": 0.0109, + "step": 99570 + }, + { + "epoch": 28.972941518766365, + "grad_norm": 0.142991840839386, + "learning_rate": 2.313200436018786e-06, + "loss": 0.0109, + "step": 99580 + }, + { + "epoch": 28.97585103287751, + "grad_norm": 0.12948636710643768, + "learning_rate": 2.3109955144232768e-06, + "loss": 0.0093, + "step": 99590 + }, + { + "epoch": 28.978760546988653, + "grad_norm": 0.09271764010190964, + "learning_rate": 2.308791556473921e-06, + "loss": 0.0104, + "step": 99600 + }, + { + "epoch": 28.981670061099795, + "grad_norm": 0.10465230792760849, + "learning_rate": 2.306588562338094e-06, + "loss": 0.0137, + "step": 99610 + }, + { + "epoch": 28.98457957521094, + "grad_norm": 0.14591771364212036, + "learning_rate": 2.304386532183092e-06, + "loss": 0.0142, + "step": 99620 + }, + { + "epoch": 28.987489089322082, + "grad_norm": 0.09760881960391998, + "learning_rate": 2.3021854661761534e-06, + "loss": 0.0125, + "step": 99630 + }, + { + "epoch": 28.990398603433228, + "grad_norm": 0.11471458524465561, + "learning_rate": 2.299985364484427e-06, + "loss": 0.0095, + "step": 99640 + }, + { + "epoch": 28.99330811754437, + "grad_norm": 0.09893053770065308, + "learning_rate": 2.297786227275003e-06, + "loss": 0.0102, + "step": 99650 + }, + { + "epoch": 28.996217631655515, + "grad_norm": 0.12702278792858124, + "learning_rate": 2.2955880547148833e-06, + "loss": 0.0111, + "step": 99660 + }, + { + "epoch": 28.999127145766657, + "grad_norm": 0.14084993302822113, + "learning_rate": 2.29339084697101e-06, + "loss": 0.0098, + "step": 99670 + }, + { + "epoch": 29.0020366598778, + "grad_norm": 0.09953160583972931, + "learning_rate": 2.291194604210249e-06, + "loss": 0.0111, + "step": 99680 + }, + { + "epoch": 29.004946173988944, + "grad_norm": 0.10567528754472733, + "learning_rate": 2.2889993265993875e-06, + "loss": 0.0088, + "step": 99690 + }, + { + "epoch": 29.007855688100086, + "grad_norm": 0.11675871163606644, + "learning_rate": 2.2868050143051343e-06, + "loss": 0.0114, + "step": 99700 + }, + { + "epoch": 29.010765202211232, + "grad_norm": 0.09852224588394165, + "learning_rate": 2.2846116674941465e-06, + "loss": 0.0115, + "step": 99710 + }, + { + "epoch": 29.013674716322374, + "grad_norm": 0.09203079342842102, + "learning_rate": 2.28241928633299e-06, + "loss": 0.0105, + "step": 99720 + }, + { + "epoch": 29.01658423043352, + "grad_norm": 0.12082663923501968, + "learning_rate": 2.2802278709881557e-06, + "loss": 0.0122, + "step": 99730 + }, + { + "epoch": 29.01949374454466, + "grad_norm": 0.10222533345222473, + "learning_rate": 2.27803742162607e-06, + "loss": 0.0091, + "step": 99740 + }, + { + "epoch": 29.022403258655803, + "grad_norm": 0.11994101852178574, + "learning_rate": 2.2758479384130903e-06, + "loss": 0.0106, + "step": 99750 + }, + { + "epoch": 29.02531277276695, + "grad_norm": 0.10856207460165024, + "learning_rate": 2.2736594215154865e-06, + "loss": 0.0092, + "step": 99760 + }, + { + "epoch": 29.02822228687809, + "grad_norm": 0.08752094954252243, + "learning_rate": 2.2714718710994594e-06, + "loss": 0.0115, + "step": 99770 + }, + { + "epoch": 29.031131800989236, + "grad_norm": 0.1148502454161644, + "learning_rate": 2.269285287331144e-06, + "loss": 0.0104, + "step": 99780 + }, + { + "epoch": 29.034041315100378, + "grad_norm": 0.11640161275863647, + "learning_rate": 2.2670996703765973e-06, + "loss": 0.0103, + "step": 99790 + }, + { + "epoch": 29.03695082921152, + "grad_norm": 0.18910947442054749, + "learning_rate": 2.264915020401799e-06, + "loss": 0.0111, + "step": 99800 + }, + { + "epoch": 29.039860343322665, + "grad_norm": 0.143405944108963, + "learning_rate": 2.2627313375726587e-06, + "loss": 0.0121, + "step": 99810 + }, + { + "epoch": 29.042769857433807, + "grad_norm": 0.10507558286190033, + "learning_rate": 2.2605486220550196e-06, + "loss": 0.0098, + "step": 99820 + }, + { + "epoch": 29.045679371544953, + "grad_norm": 0.249084934592247, + "learning_rate": 2.258366874014639e-06, + "loss": 0.0103, + "step": 99830 + }, + { + "epoch": 29.048588885656095, + "grad_norm": 0.09517933428287506, + "learning_rate": 2.256186093617204e-06, + "loss": 0.0128, + "step": 99840 + }, + { + "epoch": 29.05149839976724, + "grad_norm": 0.08562472462654114, + "learning_rate": 2.2540062810283326e-06, + "loss": 0.0106, + "step": 99850 + }, + { + "epoch": 29.054407913878382, + "grad_norm": 0.11253012716770172, + "learning_rate": 2.2518274364135716e-06, + "loss": 0.0115, + "step": 99860 + }, + { + "epoch": 29.057317427989524, + "grad_norm": 0.12640061974525452, + "learning_rate": 2.249649559938383e-06, + "loss": 0.01, + "step": 99870 + }, + { + "epoch": 29.06022694210067, + "grad_norm": 0.17042694985866547, + "learning_rate": 2.2474726517681683e-06, + "loss": 0.0095, + "step": 99880 + }, + { + "epoch": 29.06313645621181, + "grad_norm": 0.1118495836853981, + "learning_rate": 2.2452967120682444e-06, + "loss": 0.0096, + "step": 99890 + }, + { + "epoch": 29.066045970322957, + "grad_norm": 0.0998217910528183, + "learning_rate": 2.243121741003859e-06, + "loss": 0.0117, + "step": 99900 + }, + { + "epoch": 29.0689554844341, + "grad_norm": 0.08165398985147476, + "learning_rate": 2.240947738740195e-06, + "loss": 0.0096, + "step": 99910 + }, + { + "epoch": 29.071864998545244, + "grad_norm": 0.11531069874763489, + "learning_rate": 2.2387747054423436e-06, + "loss": 0.0116, + "step": 99920 + }, + { + "epoch": 29.074774512656386, + "grad_norm": 0.11632228642702103, + "learning_rate": 2.236602641275339e-06, + "loss": 0.0105, + "step": 99930 + }, + { + "epoch": 29.07768402676753, + "grad_norm": 0.13619601726531982, + "learning_rate": 2.2344315464041305e-06, + "loss": 0.0103, + "step": 99940 + }, + { + "epoch": 29.080593540878674, + "grad_norm": 0.13361766934394836, + "learning_rate": 2.232261420993603e-06, + "loss": 0.0118, + "step": 99950 + }, + { + "epoch": 29.083503054989816, + "grad_norm": 0.11400602012872696, + "learning_rate": 2.2300922652085592e-06, + "loss": 0.0106, + "step": 99960 + }, + { + "epoch": 29.08641256910096, + "grad_norm": 0.09594951570034027, + "learning_rate": 2.227924079213732e-06, + "loss": 0.0116, + "step": 99970 + }, + { + "epoch": 29.089322083212103, + "grad_norm": 0.12091630697250366, + "learning_rate": 2.2257568631737873e-06, + "loss": 0.0112, + "step": 99980 + }, + { + "epoch": 29.09223159732325, + "grad_norm": 0.11330015957355499, + "learning_rate": 2.223590617253306e-06, + "loss": 0.0118, + "step": 99990 + }, + { + "epoch": 29.09514111143439, + "grad_norm": 0.09932112693786621, + "learning_rate": 2.221425341616796e-06, + "loss": 0.0104, + "step": 100000 + }, + { + "epoch": 29.098050625545532, + "grad_norm": 0.15964160859584808, + "learning_rate": 2.2192610364286997e-06, + "loss": 0.0092, + "step": 100010 + }, + { + "epoch": 29.100960139656678, + "grad_norm": 0.11254847794771194, + "learning_rate": 2.2170977018533865e-06, + "loss": 0.0101, + "step": 100020 + }, + { + "epoch": 29.10386965376782, + "grad_norm": 0.1359476000070572, + "learning_rate": 2.214935338055141e-06, + "loss": 0.0106, + "step": 100030 + }, + { + "epoch": 29.106779167878965, + "grad_norm": 0.09331101924180984, + "learning_rate": 2.2127739451981803e-06, + "loss": 0.0103, + "step": 100040 + }, + { + "epoch": 29.109688681990107, + "grad_norm": 0.06974576413631439, + "learning_rate": 2.2106135234466545e-06, + "loss": 0.0101, + "step": 100050 + }, + { + "epoch": 29.112598196101253, + "grad_norm": 0.1043403372168541, + "learning_rate": 2.2084540729646286e-06, + "loss": 0.011, + "step": 100060 + }, + { + "epoch": 29.115507710212395, + "grad_norm": 0.27879592776298523, + "learning_rate": 2.206295593916096e-06, + "loss": 0.0103, + "step": 100070 + }, + { + "epoch": 29.118417224323537, + "grad_norm": 0.13096123933792114, + "learning_rate": 2.2041380864649814e-06, + "loss": 0.012, + "step": 100080 + }, + { + "epoch": 29.121326738434682, + "grad_norm": 0.09502067416906357, + "learning_rate": 2.2019815507751377e-06, + "loss": 0.0103, + "step": 100090 + }, + { + "epoch": 29.124236252545824, + "grad_norm": 0.12075458467006683, + "learning_rate": 2.1998259870103347e-06, + "loss": 0.0108, + "step": 100100 + }, + { + "epoch": 29.12714576665697, + "grad_norm": 0.09160366654396057, + "learning_rate": 2.1976713953342683e-06, + "loss": 0.0095, + "step": 100110 + }, + { + "epoch": 29.13005528076811, + "grad_norm": 0.09268689155578613, + "learning_rate": 2.1955177759105784e-06, + "loss": 0.0103, + "step": 100120 + }, + { + "epoch": 29.132964794879257, + "grad_norm": 0.12697972357273102, + "learning_rate": 2.193365128902811e-06, + "loss": 0.0099, + "step": 100130 + }, + { + "epoch": 29.1358743089904, + "grad_norm": 0.11594060808420181, + "learning_rate": 2.191213454474441e-06, + "loss": 0.011, + "step": 100140 + }, + { + "epoch": 29.13878382310154, + "grad_norm": 0.08746766299009323, + "learning_rate": 2.1890627527888807e-06, + "loss": 0.0112, + "step": 100150 + }, + { + "epoch": 29.141693337212686, + "grad_norm": 0.10066314786672592, + "learning_rate": 2.186913024009461e-06, + "loss": 0.0109, + "step": 100160 + }, + { + "epoch": 29.144602851323828, + "grad_norm": 0.09994985163211823, + "learning_rate": 2.184764268299436e-06, + "loss": 0.011, + "step": 100170 + }, + { + "epoch": 29.147512365434974, + "grad_norm": 0.11228153109550476, + "learning_rate": 2.1826164858219943e-06, + "loss": 0.0093, + "step": 100180 + }, + { + "epoch": 29.150421879546116, + "grad_norm": 0.08419906347990036, + "learning_rate": 2.1804696767402397e-06, + "loss": 0.0104, + "step": 100190 + }, + { + "epoch": 29.153331393657258, + "grad_norm": 0.16306298971176147, + "learning_rate": 2.1783238412172156e-06, + "loss": 0.013, + "step": 100200 + }, + { + "epoch": 29.156240907768403, + "grad_norm": 0.17592273652553558, + "learning_rate": 2.1761789794158763e-06, + "loss": 0.0142, + "step": 100210 + }, + { + "epoch": 29.159150421879545, + "grad_norm": 0.08122444897890091, + "learning_rate": 2.1740350914991135e-06, + "loss": 0.0098, + "step": 100220 + }, + { + "epoch": 29.16205993599069, + "grad_norm": 0.14341607689857483, + "learning_rate": 2.171892177629745e-06, + "loss": 0.0104, + "step": 100230 + }, + { + "epoch": 29.164969450101832, + "grad_norm": 0.08934833109378815, + "learning_rate": 2.1697502379705043e-06, + "loss": 0.0101, + "step": 100240 + }, + { + "epoch": 29.167878964212978, + "grad_norm": 0.10079644620418549, + "learning_rate": 2.167609272684064e-06, + "loss": 0.0103, + "step": 100250 + }, + { + "epoch": 29.17078847832412, + "grad_norm": 0.07793530821800232, + "learning_rate": 2.1654692819330095e-06, + "loss": 0.011, + "step": 100260 + }, + { + "epoch": 29.17369799243526, + "grad_norm": 0.11267593502998352, + "learning_rate": 2.1633302658798628e-06, + "loss": 0.0095, + "step": 100270 + }, + { + "epoch": 29.176607506546407, + "grad_norm": 0.07484762370586395, + "learning_rate": 2.16119222468707e-06, + "loss": 0.0109, + "step": 100280 + }, + { + "epoch": 29.17951702065755, + "grad_norm": 0.0997932106256485, + "learning_rate": 2.1590551585169992e-06, + "loss": 0.0094, + "step": 100290 + }, + { + "epoch": 29.182426534768695, + "grad_norm": 0.14049145579338074, + "learning_rate": 2.156919067531945e-06, + "loss": 0.0099, + "step": 100300 + }, + { + "epoch": 29.185336048879837, + "grad_norm": 0.11893592774868011, + "learning_rate": 2.154783951894129e-06, + "loss": 0.0107, + "step": 100310 + }, + { + "epoch": 29.188245562990982, + "grad_norm": 0.09182490408420563, + "learning_rate": 2.152649811765704e-06, + "loss": 0.0114, + "step": 100320 + }, + { + "epoch": 29.191155077102124, + "grad_norm": 0.0954454243183136, + "learning_rate": 2.150516647308738e-06, + "loss": 0.0094, + "step": 100330 + }, + { + "epoch": 29.194064591213266, + "grad_norm": 0.07581709325313568, + "learning_rate": 2.1483844586852325e-06, + "loss": 0.0099, + "step": 100340 + }, + { + "epoch": 29.19697410532441, + "grad_norm": 0.08594097942113876, + "learning_rate": 2.146253246057118e-06, + "loss": 0.0096, + "step": 100350 + }, + { + "epoch": 29.199883619435553, + "grad_norm": 0.06995853036642075, + "learning_rate": 2.1441230095862424e-06, + "loss": 0.0095, + "step": 100360 + }, + { + "epoch": 29.2027931335467, + "grad_norm": 0.10568541288375854, + "learning_rate": 2.141993749434379e-06, + "loss": 0.0089, + "step": 100370 + }, + { + "epoch": 29.20570264765784, + "grad_norm": 0.09679011255502701, + "learning_rate": 2.1398654657632347e-06, + "loss": 0.0095, + "step": 100380 + }, + { + "epoch": 29.208612161768986, + "grad_norm": 0.10506687313318253, + "learning_rate": 2.137738158734443e-06, + "loss": 0.0096, + "step": 100390 + }, + { + "epoch": 29.211521675880128, + "grad_norm": 0.0834222063422203, + "learning_rate": 2.135611828509554e-06, + "loss": 0.0111, + "step": 100400 + }, + { + "epoch": 29.21443118999127, + "grad_norm": 0.12911838293075562, + "learning_rate": 2.1334864752500434e-06, + "loss": 0.0114, + "step": 100410 + }, + { + "epoch": 29.217340704102416, + "grad_norm": 0.1603308469057083, + "learning_rate": 2.1313620991173288e-06, + "loss": 0.0117, + "step": 100420 + }, + { + "epoch": 29.220250218213558, + "grad_norm": 0.136185884475708, + "learning_rate": 2.1292387002727383e-06, + "loss": 0.0105, + "step": 100430 + }, + { + "epoch": 29.223159732324703, + "grad_norm": 0.07187611609697342, + "learning_rate": 2.127116278877524e-06, + "loss": 0.0107, + "step": 100440 + }, + { + "epoch": 29.226069246435845, + "grad_norm": 0.13732172548770905, + "learning_rate": 2.1249948350928755e-06, + "loss": 0.0114, + "step": 100450 + }, + { + "epoch": 29.22897876054699, + "grad_norm": 0.10140407830476761, + "learning_rate": 2.122874369079904e-06, + "loss": 0.0123, + "step": 100460 + }, + { + "epoch": 29.231888274658132, + "grad_norm": 0.13675770163536072, + "learning_rate": 2.120754880999644e-06, + "loss": 0.0114, + "step": 100470 + }, + { + "epoch": 29.234797788769274, + "grad_norm": 0.13266988098621368, + "learning_rate": 2.1186363710130518e-06, + "loss": 0.0121, + "step": 100480 + }, + { + "epoch": 29.23770730288042, + "grad_norm": 0.1064680740237236, + "learning_rate": 2.116518839281017e-06, + "loss": 0.01, + "step": 100490 + }, + { + "epoch": 29.24061681699156, + "grad_norm": 0.10683570057153702, + "learning_rate": 2.1144022859643557e-06, + "loss": 0.0102, + "step": 100500 + }, + { + "epoch": 29.243526331102707, + "grad_norm": 0.08171001076698303, + "learning_rate": 2.1122867112238016e-06, + "loss": 0.0107, + "step": 100510 + }, + { + "epoch": 29.24643584521385, + "grad_norm": 0.10762447118759155, + "learning_rate": 2.110172115220019e-06, + "loss": 0.0102, + "step": 100520 + }, + { + "epoch": 29.24934535932499, + "grad_norm": 0.10301747918128967, + "learning_rate": 2.108058498113601e-06, + "loss": 0.0109, + "step": 100530 + }, + { + "epoch": 29.252254873436137, + "grad_norm": 0.07178374379873276, + "learning_rate": 2.1059458600650593e-06, + "loss": 0.0118, + "step": 100540 + }, + { + "epoch": 29.25516438754728, + "grad_norm": 0.09938082098960876, + "learning_rate": 2.1038342012348394e-06, + "loss": 0.0109, + "step": 100550 + }, + { + "epoch": 29.258073901658424, + "grad_norm": 0.08361728489398956, + "learning_rate": 2.101723521783299e-06, + "loss": 0.0116, + "step": 100560 + }, + { + "epoch": 29.260983415769566, + "grad_norm": 0.12670157849788666, + "learning_rate": 2.099613821870742e-06, + "loss": 0.009, + "step": 100570 + }, + { + "epoch": 29.26389292988071, + "grad_norm": 0.11936090141534805, + "learning_rate": 2.0975051016573755e-06, + "loss": 0.0118, + "step": 100580 + }, + { + "epoch": 29.266802443991853, + "grad_norm": 0.09684336930513382, + "learning_rate": 2.0953973613033504e-06, + "loss": 0.0136, + "step": 100590 + }, + { + "epoch": 29.269711958102995, + "grad_norm": 0.0907241702079773, + "learning_rate": 2.0932906009687297e-06, + "loss": 0.0101, + "step": 100600 + }, + { + "epoch": 29.27262147221414, + "grad_norm": 0.09695664793252945, + "learning_rate": 2.0911848208135116e-06, + "loss": 0.0109, + "step": 100610 + }, + { + "epoch": 29.275530986325283, + "grad_norm": 0.10977193713188171, + "learning_rate": 2.0890800209976197e-06, + "loss": 0.0125, + "step": 100620 + }, + { + "epoch": 29.278440500436428, + "grad_norm": 0.09276364743709564, + "learning_rate": 2.0869762016808896e-06, + "loss": 0.0111, + "step": 100630 + }, + { + "epoch": 29.28135001454757, + "grad_norm": 0.2134445458650589, + "learning_rate": 2.0848733630231005e-06, + "loss": 0.0097, + "step": 100640 + }, + { + "epoch": 29.284259528658716, + "grad_norm": 0.0964401364326477, + "learning_rate": 2.082771505183949e-06, + "loss": 0.0085, + "step": 100650 + }, + { + "epoch": 29.287169042769857, + "grad_norm": 0.1553511768579483, + "learning_rate": 2.0806706283230563e-06, + "loss": 0.0111, + "step": 100660 + }, + { + "epoch": 29.290078556881, + "grad_norm": 0.09610084444284439, + "learning_rate": 2.0785707325999666e-06, + "loss": 0.0105, + "step": 100670 + }, + { + "epoch": 29.292988070992145, + "grad_norm": 0.08972111344337463, + "learning_rate": 2.076471818174155e-06, + "loss": 0.0107, + "step": 100680 + }, + { + "epoch": 29.295897585103287, + "grad_norm": 0.09561630338430405, + "learning_rate": 2.0743738852050243e-06, + "loss": 0.0104, + "step": 100690 + }, + { + "epoch": 29.298807099214432, + "grad_norm": 0.27085110545158386, + "learning_rate": 2.0722769338518955e-06, + "loss": 0.0099, + "step": 100700 + }, + { + "epoch": 29.301716613325574, + "grad_norm": 0.08971807360649109, + "learning_rate": 2.070180964274011e-06, + "loss": 0.0108, + "step": 100710 + }, + { + "epoch": 29.30462612743672, + "grad_norm": 0.12022259086370468, + "learning_rate": 2.068085976630559e-06, + "loss": 0.0108, + "step": 100720 + }, + { + "epoch": 29.30753564154786, + "grad_norm": 0.06886981427669525, + "learning_rate": 2.0659919710806346e-06, + "loss": 0.0099, + "step": 100730 + }, + { + "epoch": 29.310445155659004, + "grad_norm": 0.0815884917974472, + "learning_rate": 2.0638989477832598e-06, + "loss": 0.0099, + "step": 100740 + }, + { + "epoch": 29.31335466977015, + "grad_norm": 0.10361463576555252, + "learning_rate": 2.061806906897389e-06, + "loss": 0.0106, + "step": 100750 + }, + { + "epoch": 29.31626418388129, + "grad_norm": 0.11092673987150192, + "learning_rate": 2.0597158485819013e-06, + "loss": 0.0101, + "step": 100760 + }, + { + "epoch": 29.319173697992436, + "grad_norm": 0.11817178130149841, + "learning_rate": 2.057625772995598e-06, + "loss": 0.0121, + "step": 100770 + }, + { + "epoch": 29.32208321210358, + "grad_norm": 0.1493772715330124, + "learning_rate": 2.0555366802972015e-06, + "loss": 0.0115, + "step": 100780 + }, + { + "epoch": 29.324992726214724, + "grad_norm": 0.1793472021818161, + "learning_rate": 2.053448570645367e-06, + "loss": 0.0116, + "step": 100790 + }, + { + "epoch": 29.327902240325866, + "grad_norm": 0.11419256776571274, + "learning_rate": 2.0513614441986765e-06, + "loss": 0.0101, + "step": 100800 + }, + { + "epoch": 29.330811754437008, + "grad_norm": 0.092208132147789, + "learning_rate": 2.0492753011156303e-06, + "loss": 0.0104, + "step": 100810 + }, + { + "epoch": 29.333721268548153, + "grad_norm": 0.0990103930234909, + "learning_rate": 2.0471901415546556e-06, + "loss": 0.0103, + "step": 100820 + }, + { + "epoch": 29.336630782659295, + "grad_norm": 0.096556656062603, + "learning_rate": 2.0451059656741116e-06, + "loss": 0.0091, + "step": 100830 + }, + { + "epoch": 29.33954029677044, + "grad_norm": 0.13292334973812103, + "learning_rate": 2.043022773632275e-06, + "loss": 0.0104, + "step": 100840 + }, + { + "epoch": 29.342449810881583, + "grad_norm": 0.10213454067707062, + "learning_rate": 2.040940565587347e-06, + "loss": 0.0106, + "step": 100850 + }, + { + "epoch": 29.345359324992724, + "grad_norm": 0.09142401069402695, + "learning_rate": 2.0388593416974616e-06, + "loss": 0.0102, + "step": 100860 + }, + { + "epoch": 29.34826883910387, + "grad_norm": 0.14481885731220245, + "learning_rate": 2.0367791021206754e-06, + "loss": 0.0109, + "step": 100870 + }, + { + "epoch": 29.351178353215012, + "grad_norm": 0.1479727029800415, + "learning_rate": 2.0346998470149626e-06, + "loss": 0.012, + "step": 100880 + }, + { + "epoch": 29.354087867326157, + "grad_norm": 0.12052427977323532, + "learning_rate": 2.032621576538238e-06, + "loss": 0.0097, + "step": 100890 + }, + { + "epoch": 29.3569973814373, + "grad_norm": 0.14890700578689575, + "learning_rate": 2.030544290848322e-06, + "loss": 0.0116, + "step": 100900 + }, + { + "epoch": 29.359906895548445, + "grad_norm": 0.12221146374940872, + "learning_rate": 2.0284679901029766e-06, + "loss": 0.0112, + "step": 100910 + }, + { + "epoch": 29.362816409659587, + "grad_norm": 0.09359100461006165, + "learning_rate": 2.026392674459888e-06, + "loss": 0.0106, + "step": 100920 + }, + { + "epoch": 29.36572592377073, + "grad_norm": 0.10174845159053802, + "learning_rate": 2.0243183440766517e-06, + "loss": 0.0095, + "step": 100930 + }, + { + "epoch": 29.368635437881874, + "grad_norm": 0.10905132442712784, + "learning_rate": 2.0222449991108094e-06, + "loss": 0.0121, + "step": 100940 + }, + { + "epoch": 29.371544951993016, + "grad_norm": 0.10710779577493668, + "learning_rate": 2.0201726397198104e-06, + "loss": 0.0103, + "step": 100950 + }, + { + "epoch": 29.37445446610416, + "grad_norm": 0.13362687826156616, + "learning_rate": 2.0181012660610443e-06, + "loss": 0.0134, + "step": 100960 + }, + { + "epoch": 29.377363980215303, + "grad_norm": 0.0569881796836853, + "learning_rate": 2.0160308782918095e-06, + "loss": 0.0107, + "step": 100970 + }, + { + "epoch": 29.38027349432645, + "grad_norm": 0.12120698392391205, + "learning_rate": 2.0139614765693426e-06, + "loss": 0.0103, + "step": 100980 + }, + { + "epoch": 29.38318300843759, + "grad_norm": 0.11228722333908081, + "learning_rate": 2.0118930610508058e-06, + "loss": 0.0122, + "step": 100990 + }, + { + "epoch": 29.386092522548733, + "grad_norm": 0.15557563304901123, + "learning_rate": 2.009825631893275e-06, + "loss": 0.0106, + "step": 101000 + }, + { + "epoch": 29.38900203665988, + "grad_norm": 0.14742694795131683, + "learning_rate": 2.007759189253755e-06, + "loss": 0.0119, + "step": 101010 + }, + { + "epoch": 29.39191155077102, + "grad_norm": 0.09245942533016205, + "learning_rate": 2.0056937332891882e-06, + "loss": 0.0099, + "step": 101020 + }, + { + "epoch": 29.394821064882166, + "grad_norm": 0.17287573218345642, + "learning_rate": 2.003629264156428e-06, + "loss": 0.0099, + "step": 101030 + }, + { + "epoch": 29.397730578993308, + "grad_norm": 0.10807964950799942, + "learning_rate": 2.0015657820122534e-06, + "loss": 0.0094, + "step": 101040 + }, + { + "epoch": 29.400640093104453, + "grad_norm": 0.13151443004608154, + "learning_rate": 1.999503287013376e-06, + "loss": 0.01, + "step": 101050 + }, + { + "epoch": 29.403549607215595, + "grad_norm": 0.1426835060119629, + "learning_rate": 1.9974417793164305e-06, + "loss": 0.0103, + "step": 101060 + }, + { + "epoch": 29.406459121326737, + "grad_norm": 0.0971461832523346, + "learning_rate": 1.9953812590779735e-06, + "loss": 0.0129, + "step": 101070 + }, + { + "epoch": 29.409368635437882, + "grad_norm": 0.08606117218732834, + "learning_rate": 1.993321726454483e-06, + "loss": 0.0104, + "step": 101080 + }, + { + "epoch": 29.412278149549024, + "grad_norm": 0.17918094992637634, + "learning_rate": 1.9912631816023718e-06, + "loss": 0.0105, + "step": 101090 + }, + { + "epoch": 29.41518766366017, + "grad_norm": 0.0989542081952095, + "learning_rate": 1.9892056246779737e-06, + "loss": 0.0112, + "step": 101100 + }, + { + "epoch": 29.418097177771312, + "grad_norm": 0.07978363335132599, + "learning_rate": 1.9871490558375404e-06, + "loss": 0.0104, + "step": 101110 + }, + { + "epoch": 29.421006691882457, + "grad_norm": 0.10640457272529602, + "learning_rate": 1.985093475237262e-06, + "loss": 0.01, + "step": 101120 + }, + { + "epoch": 29.4239162059936, + "grad_norm": 0.07783504575490952, + "learning_rate": 1.983038883033245e-06, + "loss": 0.0099, + "step": 101130 + }, + { + "epoch": 29.42682572010474, + "grad_norm": 0.11495430767536163, + "learning_rate": 1.980985279381519e-06, + "loss": 0.0093, + "step": 101140 + }, + { + "epoch": 29.429735234215887, + "grad_norm": 0.12741905450820923, + "learning_rate": 1.9789326644380407e-06, + "loss": 0.0092, + "step": 101150 + }, + { + "epoch": 29.43264474832703, + "grad_norm": 0.11289866268634796, + "learning_rate": 1.976881038358696e-06, + "loss": 0.0104, + "step": 101160 + }, + { + "epoch": 29.435554262438174, + "grad_norm": 0.08221495896577835, + "learning_rate": 1.9748304012992925e-06, + "loss": 0.0107, + "step": 101170 + }, + { + "epoch": 29.438463776549316, + "grad_norm": 0.07474654912948608, + "learning_rate": 1.9727807534155602e-06, + "loss": 0.0108, + "step": 101180 + }, + { + "epoch": 29.441373290660458, + "grad_norm": 0.09814850986003876, + "learning_rate": 1.97073209486316e-06, + "loss": 0.0112, + "step": 101190 + }, + { + "epoch": 29.444282804771603, + "grad_norm": 0.10816729813814163, + "learning_rate": 1.9686844257976683e-06, + "loss": 0.0104, + "step": 101200 + }, + { + "epoch": 29.447192318882745, + "grad_norm": 0.09502719342708588, + "learning_rate": 1.966637746374598e-06, + "loss": 0.0125, + "step": 101210 + }, + { + "epoch": 29.45010183299389, + "grad_norm": 0.09641389548778534, + "learning_rate": 1.9645920567493753e-06, + "loss": 0.0103, + "step": 101220 + }, + { + "epoch": 29.453011347105033, + "grad_norm": 0.09855595231056213, + "learning_rate": 1.96254735707736e-06, + "loss": 0.0094, + "step": 101230 + }, + { + "epoch": 29.45592086121618, + "grad_norm": 0.06335611641407013, + "learning_rate": 1.9605036475138368e-06, + "loss": 0.0112, + "step": 101240 + }, + { + "epoch": 29.45883037532732, + "grad_norm": 0.1334671974182129, + "learning_rate": 1.9584609282140043e-06, + "loss": 0.0112, + "step": 101250 + }, + { + "epoch": 29.461739889438462, + "grad_norm": 0.11383145302534103, + "learning_rate": 1.9564191993330003e-06, + "loss": 0.0116, + "step": 101260 + }, + { + "epoch": 29.464649403549608, + "grad_norm": 0.14884626865386963, + "learning_rate": 1.9543784610258764e-06, + "loss": 0.011, + "step": 101270 + }, + { + "epoch": 29.46755891766075, + "grad_norm": 0.10305126011371613, + "learning_rate": 1.952338713447613e-06, + "loss": 0.0109, + "step": 101280 + }, + { + "epoch": 29.470468431771895, + "grad_norm": 0.1396833211183548, + "learning_rate": 1.950299956753121e-06, + "loss": 0.0123, + "step": 101290 + }, + { + "epoch": 29.473377945883037, + "grad_norm": 0.09662694483995438, + "learning_rate": 1.948262191097227e-06, + "loss": 0.0115, + "step": 101300 + }, + { + "epoch": 29.476287459994182, + "grad_norm": 0.10248537361621857, + "learning_rate": 1.946225416634684e-06, + "loss": 0.0112, + "step": 101310 + }, + { + "epoch": 29.479196974105324, + "grad_norm": 0.08456873893737793, + "learning_rate": 1.944189633520171e-06, + "loss": 0.0115, + "step": 101320 + }, + { + "epoch": 29.482106488216466, + "grad_norm": 0.07734515517950058, + "learning_rate": 1.9421548419082978e-06, + "loss": 0.0101, + "step": 101330 + }, + { + "epoch": 29.48501600232761, + "grad_norm": 0.1411316692829132, + "learning_rate": 1.9401210419535883e-06, + "loss": 0.0104, + "step": 101340 + }, + { + "epoch": 29.487925516438754, + "grad_norm": 0.111363984644413, + "learning_rate": 1.938088233810496e-06, + "loss": 0.0091, + "step": 101350 + }, + { + "epoch": 29.4908350305499, + "grad_norm": 0.07467706501483917, + "learning_rate": 1.9360564176334068e-06, + "loss": 0.0104, + "step": 101360 + }, + { + "epoch": 29.49374454466104, + "grad_norm": 0.0759497657418251, + "learning_rate": 1.9340255935766164e-06, + "loss": 0.0106, + "step": 101370 + }, + { + "epoch": 29.496654058772187, + "grad_norm": 0.1001879870891571, + "learning_rate": 1.9319957617943505e-06, + "loss": 0.0093, + "step": 101380 + }, + { + "epoch": 29.49956357288333, + "grad_norm": 0.11414728313684464, + "learning_rate": 1.9299669224407674e-06, + "loss": 0.0078, + "step": 101390 + }, + { + "epoch": 29.50247308699447, + "grad_norm": 0.11445130407810211, + "learning_rate": 1.9279390756699435e-06, + "loss": 0.012, + "step": 101400 + }, + { + "epoch": 29.505382601105616, + "grad_norm": 0.1151626780629158, + "learning_rate": 1.925912221635879e-06, + "loss": 0.009, + "step": 101410 + }, + { + "epoch": 29.508292115216758, + "grad_norm": 0.10042371600866318, + "learning_rate": 1.9238863604924937e-06, + "loss": 0.0103, + "step": 101420 + }, + { + "epoch": 29.511201629327903, + "grad_norm": 0.11989793181419373, + "learning_rate": 1.9218614923936508e-06, + "loss": 0.0094, + "step": 101430 + }, + { + "epoch": 29.514111143439045, + "grad_norm": 0.15080603957176208, + "learning_rate": 1.919837617493119e-06, + "loss": 0.0108, + "step": 101440 + }, + { + "epoch": 29.51702065755019, + "grad_norm": 0.08879242092370987, + "learning_rate": 1.9178147359445953e-06, + "loss": 0.0117, + "step": 101450 + }, + { + "epoch": 29.519930171661333, + "grad_norm": 0.14672519266605377, + "learning_rate": 1.915792847901707e-06, + "loss": 0.0108, + "step": 101460 + }, + { + "epoch": 29.522839685772475, + "grad_norm": 0.12680429220199585, + "learning_rate": 1.9137719535180077e-06, + "loss": 0.0113, + "step": 101470 + }, + { + "epoch": 29.52574919988362, + "grad_norm": 0.11752855777740479, + "learning_rate": 1.9117520529469652e-06, + "loss": 0.0104, + "step": 101480 + }, + { + "epoch": 29.528658713994762, + "grad_norm": 0.12188045680522919, + "learning_rate": 1.9097331463419805e-06, + "loss": 0.0105, + "step": 101490 + }, + { + "epoch": 29.531568228105908, + "grad_norm": 0.05813829228281975, + "learning_rate": 1.9077152338563724e-06, + "loss": 0.0104, + "step": 101500 + }, + { + "epoch": 29.53447774221705, + "grad_norm": 0.09863369166851044, + "learning_rate": 1.9056983156433939e-06, + "loss": 0.0099, + "step": 101510 + }, + { + "epoch": 29.537387256328195, + "grad_norm": 0.12428983300924301, + "learning_rate": 1.9036823918562102e-06, + "loss": 0.0125, + "step": 101520 + }, + { + "epoch": 29.540296770439337, + "grad_norm": 0.1044844314455986, + "learning_rate": 1.9016674626479208e-06, + "loss": 0.0117, + "step": 101530 + }, + { + "epoch": 29.54320628455048, + "grad_norm": 0.12420742958784103, + "learning_rate": 1.8996535281715483e-06, + "loss": 0.0105, + "step": 101540 + }, + { + "epoch": 29.546115798661624, + "grad_norm": 0.13989852368831635, + "learning_rate": 1.8976405885800335e-06, + "loss": 0.0107, + "step": 101550 + }, + { + "epoch": 29.549025312772766, + "grad_norm": 0.10063424706459045, + "learning_rate": 1.895628644026251e-06, + "loss": 0.0106, + "step": 101560 + }, + { + "epoch": 29.55193482688391, + "grad_norm": 0.08218695968389511, + "learning_rate": 1.8936176946629886e-06, + "loss": 0.0098, + "step": 101570 + }, + { + "epoch": 29.554844340995054, + "grad_norm": 0.1434507519006729, + "learning_rate": 1.8916077406429706e-06, + "loss": 0.0124, + "step": 101580 + }, + { + "epoch": 29.557753855106196, + "grad_norm": 0.10283982753753662, + "learning_rate": 1.8895987821188348e-06, + "loss": 0.01, + "step": 101590 + }, + { + "epoch": 29.56066336921734, + "grad_norm": 0.07965666800737381, + "learning_rate": 1.8875908192431512e-06, + "loss": 0.0101, + "step": 101600 + }, + { + "epoch": 29.563572883328483, + "grad_norm": 0.13167135417461395, + "learning_rate": 1.8855838521684093e-06, + "loss": 0.01, + "step": 101610 + }, + { + "epoch": 29.56648239743963, + "grad_norm": 0.110337033867836, + "learning_rate": 1.8835778810470255e-06, + "loss": 0.0112, + "step": 101620 + }, + { + "epoch": 29.56939191155077, + "grad_norm": 0.09886447340250015, + "learning_rate": 1.8815729060313448e-06, + "loss": 0.0099, + "step": 101630 + }, + { + "epoch": 29.572301425661916, + "grad_norm": 0.11939137428998947, + "learning_rate": 1.8795689272736232e-06, + "loss": 0.0108, + "step": 101640 + }, + { + "epoch": 29.575210939773058, + "grad_norm": 0.08539621531963348, + "learning_rate": 1.8775659449260545e-06, + "loss": 0.0129, + "step": 101650 + }, + { + "epoch": 29.5781204538842, + "grad_norm": 0.10148588567972183, + "learning_rate": 1.8755639591407554e-06, + "loss": 0.0093, + "step": 101660 + }, + { + "epoch": 29.581029967995345, + "grad_norm": 0.11697600036859512, + "learning_rate": 1.8735629700697604e-06, + "loss": 0.0097, + "step": 101670 + }, + { + "epoch": 29.583939482106487, + "grad_norm": 0.08920827507972717, + "learning_rate": 1.8715629778650267e-06, + "loss": 0.0103, + "step": 101680 + }, + { + "epoch": 29.586848996217633, + "grad_norm": 0.13706074655056, + "learning_rate": 1.8695639826784461e-06, + "loss": 0.0123, + "step": 101690 + }, + { + "epoch": 29.589758510328775, + "grad_norm": 0.15870434045791626, + "learning_rate": 1.8675659846618287e-06, + "loss": 0.0112, + "step": 101700 + }, + { + "epoch": 29.59266802443992, + "grad_norm": 0.08806022256612778, + "learning_rate": 1.86556898396691e-06, + "loss": 0.0094, + "step": 101710 + }, + { + "epoch": 29.595577538551062, + "grad_norm": 0.11320655792951584, + "learning_rate": 1.86357298074534e-06, + "loss": 0.0129, + "step": 101720 + }, + { + "epoch": 29.598487052662204, + "grad_norm": 0.28692761063575745, + "learning_rate": 1.861577975148716e-06, + "loss": 0.0089, + "step": 101730 + }, + { + "epoch": 29.60139656677335, + "grad_norm": 0.11090050637722015, + "learning_rate": 1.8595839673285386e-06, + "loss": 0.0118, + "step": 101740 + }, + { + "epoch": 29.60430608088449, + "grad_norm": 0.13300001621246338, + "learning_rate": 1.8575909574362365e-06, + "loss": 0.0115, + "step": 101750 + }, + { + "epoch": 29.607215594995637, + "grad_norm": 0.07972249388694763, + "learning_rate": 1.8555989456231687e-06, + "loss": 0.0123, + "step": 101760 + }, + { + "epoch": 29.61012510910678, + "grad_norm": 0.14499983191490173, + "learning_rate": 1.8536079320406191e-06, + "loss": 0.0111, + "step": 101770 + }, + { + "epoch": 29.613034623217924, + "grad_norm": 0.11848033219575882, + "learning_rate": 1.8516179168397885e-06, + "loss": 0.0102, + "step": 101780 + }, + { + "epoch": 29.615944137329066, + "grad_norm": 0.10531267523765564, + "learning_rate": 1.8496289001718025e-06, + "loss": 0.0103, + "step": 101790 + }, + { + "epoch": 29.618853651440208, + "grad_norm": 0.08717234432697296, + "learning_rate": 1.8476408821877156e-06, + "loss": 0.0131, + "step": 101800 + }, + { + "epoch": 29.621763165551354, + "grad_norm": 0.12764336168766022, + "learning_rate": 1.8456538630385084e-06, + "loss": 0.0112, + "step": 101810 + }, + { + "epoch": 29.624672679662496, + "grad_norm": 0.14781703054904938, + "learning_rate": 1.843667842875077e-06, + "loss": 0.011, + "step": 101820 + }, + { + "epoch": 29.62758219377364, + "grad_norm": 0.12605774402618408, + "learning_rate": 1.841682821848249e-06, + "loss": 0.0114, + "step": 101830 + }, + { + "epoch": 29.630491707884783, + "grad_norm": 0.09296493977308273, + "learning_rate": 1.8396988001087756e-06, + "loss": 0.0126, + "step": 101840 + }, + { + "epoch": 29.633401221995925, + "grad_norm": 0.10579824447631836, + "learning_rate": 1.8377157778073261e-06, + "loss": 0.0096, + "step": 101850 + }, + { + "epoch": 29.63631073610707, + "grad_norm": 0.10037817060947418, + "learning_rate": 1.835733755094502e-06, + "loss": 0.0095, + "step": 101860 + }, + { + "epoch": 29.639220250218212, + "grad_norm": 0.07390552014112473, + "learning_rate": 1.8337527321208193e-06, + "loss": 0.0138, + "step": 101870 + }, + { + "epoch": 29.642129764329358, + "grad_norm": 0.1700415462255478, + "learning_rate": 1.8317727090367292e-06, + "loss": 0.0091, + "step": 101880 + }, + { + "epoch": 29.6450392784405, + "grad_norm": 0.10085412859916687, + "learning_rate": 1.8297936859925984e-06, + "loss": 0.0108, + "step": 101890 + }, + { + "epoch": 29.647948792551645, + "grad_norm": 0.11723766475915909, + "learning_rate": 1.8278156631387216e-06, + "loss": 0.012, + "step": 101900 + }, + { + "epoch": 29.650858306662787, + "grad_norm": 0.1000712588429451, + "learning_rate": 1.8258386406253152e-06, + "loss": 0.0095, + "step": 101910 + }, + { + "epoch": 29.65376782077393, + "grad_norm": 0.12283973395824432, + "learning_rate": 1.8238626186025226e-06, + "loss": 0.0112, + "step": 101920 + }, + { + "epoch": 29.656677334885075, + "grad_norm": 0.10673637688159943, + "learning_rate": 1.8218875972204101e-06, + "loss": 0.0097, + "step": 101930 + }, + { + "epoch": 29.659586848996216, + "grad_norm": 0.12367621809244156, + "learning_rate": 1.8199135766289648e-06, + "loss": 0.0103, + "step": 101940 + }, + { + "epoch": 29.662496363107362, + "grad_norm": 0.08036265522241592, + "learning_rate": 1.8179405569781032e-06, + "loss": 0.0105, + "step": 101950 + }, + { + "epoch": 29.665405877218504, + "grad_norm": 0.118735671043396, + "learning_rate": 1.8159685384176638e-06, + "loss": 0.0104, + "step": 101960 + }, + { + "epoch": 29.66831539132965, + "grad_norm": 0.1949208378791809, + "learning_rate": 1.813997521097407e-06, + "loss": 0.0119, + "step": 101970 + }, + { + "epoch": 29.67122490544079, + "grad_norm": 0.12386653572320938, + "learning_rate": 1.812027505167016e-06, + "loss": 0.0114, + "step": 101980 + }, + { + "epoch": 29.674134419551933, + "grad_norm": 0.1726614385843277, + "learning_rate": 1.810058490776103e-06, + "loss": 0.0114, + "step": 101990 + }, + { + "epoch": 29.67704393366308, + "grad_norm": 0.08835361897945404, + "learning_rate": 1.8080904780742052e-06, + "loss": 0.0106, + "step": 102000 + }, + { + "epoch": 29.67995344777422, + "grad_norm": 0.09935850650072098, + "learning_rate": 1.8061234672107747e-06, + "loss": 0.0119, + "step": 102010 + }, + { + "epoch": 29.682862961885366, + "grad_norm": 0.1313389390707016, + "learning_rate": 1.80415745833519e-06, + "loss": 0.0099, + "step": 102020 + }, + { + "epoch": 29.685772475996508, + "grad_norm": 0.15757817029953003, + "learning_rate": 1.802192451596767e-06, + "loss": 0.0112, + "step": 102030 + }, + { + "epoch": 29.688681990107654, + "grad_norm": 0.08893778175115585, + "learning_rate": 1.8002284471447311e-06, + "loss": 0.0143, + "step": 102040 + }, + { + "epoch": 29.691591504218795, + "grad_norm": 0.0861179456114769, + "learning_rate": 1.7982654451282282e-06, + "loss": 0.0095, + "step": 102050 + }, + { + "epoch": 29.694501018329937, + "grad_norm": 0.0965023934841156, + "learning_rate": 1.7963034456963424e-06, + "loss": 0.0103, + "step": 102060 + }, + { + "epoch": 29.697410532441083, + "grad_norm": 0.1216684877872467, + "learning_rate": 1.7943424489980758e-06, + "loss": 0.0128, + "step": 102070 + }, + { + "epoch": 29.700320046552225, + "grad_norm": 0.09902364760637283, + "learning_rate": 1.7923824551823499e-06, + "loss": 0.01, + "step": 102080 + }, + { + "epoch": 29.70322956066337, + "grad_norm": 0.11828017234802246, + "learning_rate": 1.7904234643980117e-06, + "loss": 0.0116, + "step": 102090 + }, + { + "epoch": 29.706139074774512, + "grad_norm": 0.11265537142753601, + "learning_rate": 1.788465476793834e-06, + "loss": 0.0108, + "step": 102100 + }, + { + "epoch": 29.709048588885658, + "grad_norm": 0.08566176146268845, + "learning_rate": 1.78650849251852e-06, + "loss": 0.0121, + "step": 102110 + }, + { + "epoch": 29.7119581029968, + "grad_norm": 0.27605873346328735, + "learning_rate": 1.7845525117206802e-06, + "loss": 0.0114, + "step": 102120 + }, + { + "epoch": 29.71486761710794, + "grad_norm": 0.07160203158855438, + "learning_rate": 1.7825975345488626e-06, + "loss": 0.0102, + "step": 102130 + }, + { + "epoch": 29.717777131219087, + "grad_norm": 0.11537927389144897, + "learning_rate": 1.7806435611515387e-06, + "loss": 0.0109, + "step": 102140 + }, + { + "epoch": 29.72068664533023, + "grad_norm": 0.18365629017353058, + "learning_rate": 1.7786905916770962e-06, + "loss": 0.01, + "step": 102150 + }, + { + "epoch": 29.723596159441374, + "grad_norm": 0.11191767454147339, + "learning_rate": 1.7767386262738483e-06, + "loss": 0.0104, + "step": 102160 + }, + { + "epoch": 29.726505673552516, + "grad_norm": 0.1435288041830063, + "learning_rate": 1.774787665090033e-06, + "loss": 0.0098, + "step": 102170 + }, + { + "epoch": 29.729415187663662, + "grad_norm": 0.09715932607650757, + "learning_rate": 1.7728377082738218e-06, + "loss": 0.0106, + "step": 102180 + }, + { + "epoch": 29.732324701774804, + "grad_norm": 0.15426063537597656, + "learning_rate": 1.7708887559732894e-06, + "loss": 0.0099, + "step": 102190 + }, + { + "epoch": 29.735234215885946, + "grad_norm": 0.09964562952518463, + "learning_rate": 1.768940808336456e-06, + "loss": 0.0099, + "step": 102200 + }, + { + "epoch": 29.73814372999709, + "grad_norm": 0.12124449759721756, + "learning_rate": 1.7669938655112467e-06, + "loss": 0.0105, + "step": 102210 + }, + { + "epoch": 29.741053244108233, + "grad_norm": 0.1147916316986084, + "learning_rate": 1.7650479276455244e-06, + "loss": 0.0099, + "step": 102220 + }, + { + "epoch": 29.74396275821938, + "grad_norm": 0.14069874584674835, + "learning_rate": 1.7631029948870713e-06, + "loss": 0.0118, + "step": 102230 + }, + { + "epoch": 29.74687227233052, + "grad_norm": 0.12165278196334839, + "learning_rate": 1.7611590673835859e-06, + "loss": 0.0082, + "step": 102240 + }, + { + "epoch": 29.749781786441666, + "grad_norm": 0.09283885359764099, + "learning_rate": 1.759216145282705e-06, + "loss": 0.0115, + "step": 102250 + }, + { + "epoch": 29.752691300552808, + "grad_norm": 0.06405047327280045, + "learning_rate": 1.757274228731972e-06, + "loss": 0.0097, + "step": 102260 + }, + { + "epoch": 29.75560081466395, + "grad_norm": 0.14725351333618164, + "learning_rate": 1.755333317878871e-06, + "loss": 0.0096, + "step": 102270 + }, + { + "epoch": 29.758510328775095, + "grad_norm": 0.12912625074386597, + "learning_rate": 1.753393412870794e-06, + "loss": 0.0114, + "step": 102280 + }, + { + "epoch": 29.761419842886237, + "grad_norm": 0.07354626804590225, + "learning_rate": 1.751454513855068e-06, + "loss": 0.0104, + "step": 102290 + }, + { + "epoch": 29.764329356997383, + "grad_norm": 0.10961515456438065, + "learning_rate": 1.7495166209789404e-06, + "loss": 0.0106, + "step": 102300 + }, + { + "epoch": 29.767238871108525, + "grad_norm": 0.08846455812454224, + "learning_rate": 1.74757973438958e-06, + "loss": 0.0095, + "step": 102310 + }, + { + "epoch": 29.770148385219667, + "grad_norm": 0.09661451727151871, + "learning_rate": 1.7456438542340758e-06, + "loss": 0.0097, + "step": 102320 + }, + { + "epoch": 29.773057899330812, + "grad_norm": 0.20122584700584412, + "learning_rate": 1.7437089806594553e-06, + "loss": 0.0103, + "step": 102330 + }, + { + "epoch": 29.775967413441954, + "grad_norm": 0.050072576850652695, + "learning_rate": 1.7417751138126525e-06, + "loss": 0.0103, + "step": 102340 + }, + { + "epoch": 29.7788769275531, + "grad_norm": 0.10244257003068924, + "learning_rate": 1.7398422538405317e-06, + "loss": 0.0084, + "step": 102350 + }, + { + "epoch": 29.78178644166424, + "grad_norm": 0.1100853830575943, + "learning_rate": 1.7379104008898805e-06, + "loss": 0.0107, + "step": 102360 + }, + { + "epoch": 29.784695955775387, + "grad_norm": 0.1796657145023346, + "learning_rate": 1.7359795551074165e-06, + "loss": 0.01, + "step": 102370 + }, + { + "epoch": 29.78760546988653, + "grad_norm": 0.12397977709770203, + "learning_rate": 1.7340497166397707e-06, + "loss": 0.0113, + "step": 102380 + }, + { + "epoch": 29.79051498399767, + "grad_norm": 0.08723917603492737, + "learning_rate": 1.7321208856334974e-06, + "loss": 0.0112, + "step": 102390 + }, + { + "epoch": 29.793424498108816, + "grad_norm": 0.14286616444587708, + "learning_rate": 1.7301930622350832e-06, + "loss": 0.0091, + "step": 102400 + }, + { + "epoch": 29.79633401221996, + "grad_norm": 0.06366188824176788, + "learning_rate": 1.7282662465909355e-06, + "loss": 0.0092, + "step": 102410 + }, + { + "epoch": 29.799243526331104, + "grad_norm": 0.13555972278118134, + "learning_rate": 1.7263404388473807e-06, + "loss": 0.0109, + "step": 102420 + }, + { + "epoch": 29.802153040442246, + "grad_norm": 0.23167648911476135, + "learning_rate": 1.7244156391506654e-06, + "loss": 0.0101, + "step": 102430 + }, + { + "epoch": 29.80506255455339, + "grad_norm": 0.0960502177476883, + "learning_rate": 1.7224918476469758e-06, + "loss": 0.0101, + "step": 102440 + }, + { + "epoch": 29.807972068664533, + "grad_norm": 0.1119316816329956, + "learning_rate": 1.7205690644824068e-06, + "loss": 0.0123, + "step": 102450 + }, + { + "epoch": 29.810881582775675, + "grad_norm": 0.12297827750444412, + "learning_rate": 1.7186472898029782e-06, + "loss": 0.0102, + "step": 102460 + }, + { + "epoch": 29.81379109688682, + "grad_norm": 0.12361941486597061, + "learning_rate": 1.7167265237546382e-06, + "loss": 0.0128, + "step": 102470 + }, + { + "epoch": 29.816700610997962, + "grad_norm": 0.11140082031488419, + "learning_rate": 1.71480676648326e-06, + "loss": 0.0117, + "step": 102480 + }, + { + "epoch": 29.819610125109108, + "grad_norm": 0.11519839614629745, + "learning_rate": 1.7128880181346307e-06, + "loss": 0.0151, + "step": 102490 + }, + { + "epoch": 29.82251963922025, + "grad_norm": 0.16340799629688263, + "learning_rate": 1.7109702788544718e-06, + "loss": 0.0115, + "step": 102500 + }, + { + "epoch": 29.825429153331395, + "grad_norm": 0.13667264580726624, + "learning_rate": 1.7090535487884167e-06, + "loss": 0.0103, + "step": 102510 + }, + { + "epoch": 29.828338667442537, + "grad_norm": 0.0913395956158638, + "learning_rate": 1.707137828082036e-06, + "loss": 0.012, + "step": 102520 + }, + { + "epoch": 29.83124818155368, + "grad_norm": 0.15245436131954193, + "learning_rate": 1.705223116880808e-06, + "loss": 0.0102, + "step": 102530 + }, + { + "epoch": 29.834157695664825, + "grad_norm": 0.1046397015452385, + "learning_rate": 1.7033094153301466e-06, + "loss": 0.011, + "step": 102540 + }, + { + "epoch": 29.837067209775967, + "grad_norm": 0.08179458230733871, + "learning_rate": 1.7013967235753886e-06, + "loss": 0.0113, + "step": 102550 + }, + { + "epoch": 29.839976723887112, + "grad_norm": 0.13974329829216003, + "learning_rate": 1.699485041761783e-06, + "loss": 0.0101, + "step": 102560 + }, + { + "epoch": 29.842886237998254, + "grad_norm": 0.1355580985546112, + "learning_rate": 1.6975743700345154e-06, + "loss": 0.0099, + "step": 102570 + }, + { + "epoch": 29.845795752109396, + "grad_norm": 0.08333700150251389, + "learning_rate": 1.695664708538683e-06, + "loss": 0.0115, + "step": 102580 + }, + { + "epoch": 29.84870526622054, + "grad_norm": 0.16347995400428772, + "learning_rate": 1.6937560574193145e-06, + "loss": 0.0107, + "step": 102590 + }, + { + "epoch": 29.851614780331683, + "grad_norm": 0.07840604335069656, + "learning_rate": 1.6918484168213627e-06, + "loss": 0.0126, + "step": 102600 + }, + { + "epoch": 29.85452429444283, + "grad_norm": 0.09157735109329224, + "learning_rate": 1.6899417868896983e-06, + "loss": 0.0128, + "step": 102610 + }, + { + "epoch": 29.85743380855397, + "grad_norm": 0.1562497913837433, + "learning_rate": 1.688036167769112e-06, + "loss": 0.0111, + "step": 102620 + }, + { + "epoch": 29.860343322665116, + "grad_norm": 0.09241738170385361, + "learning_rate": 1.6861315596043297e-06, + "loss": 0.0111, + "step": 102630 + }, + { + "epoch": 29.863252836776258, + "grad_norm": 0.08529426902532578, + "learning_rate": 1.6842279625399925e-06, + "loss": 0.012, + "step": 102640 + }, + { + "epoch": 29.8661623508874, + "grad_norm": 0.11504929512739182, + "learning_rate": 1.6823253767206632e-06, + "loss": 0.0106, + "step": 102650 + }, + { + "epoch": 29.869071864998546, + "grad_norm": 0.14017443358898163, + "learning_rate": 1.6804238022908326e-06, + "loss": 0.0132, + "step": 102660 + }, + { + "epoch": 29.871981379109688, + "grad_norm": 0.10535626113414764, + "learning_rate": 1.6785232393949152e-06, + "loss": 0.0095, + "step": 102670 + }, + { + "epoch": 29.874890893220833, + "grad_norm": 0.09237831085920334, + "learning_rate": 1.6766236881772439e-06, + "loss": 0.0107, + "step": 102680 + }, + { + "epoch": 29.877800407331975, + "grad_norm": 0.1218201294541359, + "learning_rate": 1.6747251487820736e-06, + "loss": 0.0087, + "step": 102690 + }, + { + "epoch": 29.88070992144312, + "grad_norm": 0.18164633214473724, + "learning_rate": 1.6728276213535903e-06, + "loss": 0.0131, + "step": 102700 + }, + { + "epoch": 29.883619435554262, + "grad_norm": 0.09517299383878708, + "learning_rate": 1.6709311060358989e-06, + "loss": 0.01, + "step": 102710 + }, + { + "epoch": 29.886528949665404, + "grad_norm": 0.06624863296747208, + "learning_rate": 1.6690356029730274e-06, + "loss": 0.0096, + "step": 102720 + }, + { + "epoch": 29.88943846377655, + "grad_norm": 0.09960609674453735, + "learning_rate": 1.6671411123089175e-06, + "loss": 0.0117, + "step": 102730 + }, + { + "epoch": 29.89234797788769, + "grad_norm": 0.12041423469781876, + "learning_rate": 1.6652476341874572e-06, + "loss": 0.0115, + "step": 102740 + }, + { + "epoch": 29.895257491998837, + "grad_norm": 0.12794938683509827, + "learning_rate": 1.6633551687524385e-06, + "loss": 0.0106, + "step": 102750 + }, + { + "epoch": 29.89816700610998, + "grad_norm": 0.250029593706131, + "learning_rate": 1.661463716147576e-06, + "loss": 0.011, + "step": 102760 + }, + { + "epoch": 29.901076520221125, + "grad_norm": 0.1258418709039688, + "learning_rate": 1.6595732765165184e-06, + "loss": 0.0135, + "step": 102770 + }, + { + "epoch": 29.903986034332267, + "grad_norm": 0.06732446700334549, + "learning_rate": 1.657683850002834e-06, + "loss": 0.0095, + "step": 102780 + }, + { + "epoch": 29.90689554844341, + "grad_norm": 0.2555878758430481, + "learning_rate": 1.65579543675001e-06, + "loss": 0.0108, + "step": 102790 + }, + { + "epoch": 29.909805062554554, + "grad_norm": 0.09978805482387543, + "learning_rate": 1.6539080369014564e-06, + "loss": 0.0121, + "step": 102800 + }, + { + "epoch": 29.912714576665696, + "grad_norm": 0.07008997350931168, + "learning_rate": 1.6520216506005087e-06, + "loss": 0.0087, + "step": 102810 + }, + { + "epoch": 29.91562409077684, + "grad_norm": 0.09952642023563385, + "learning_rate": 1.650136277990432e-06, + "loss": 0.0101, + "step": 102820 + }, + { + "epoch": 29.918533604887983, + "grad_norm": 0.10367520153522491, + "learning_rate": 1.6482519192144007e-06, + "loss": 0.0084, + "step": 102830 + }, + { + "epoch": 29.92144311899913, + "grad_norm": 0.0878778025507927, + "learning_rate": 1.6463685744155215e-06, + "loss": 0.0103, + "step": 102840 + }, + { + "epoch": 29.92435263311027, + "grad_norm": 0.13769274950027466, + "learning_rate": 1.6444862437368268e-06, + "loss": 0.0103, + "step": 102850 + }, + { + "epoch": 29.927262147221413, + "grad_norm": 0.10060884803533554, + "learning_rate": 1.642604927321259e-06, + "loss": 0.0105, + "step": 102860 + }, + { + "epoch": 29.930171661332558, + "grad_norm": 0.12200016528367996, + "learning_rate": 1.6407246253116987e-06, + "loss": 0.0111, + "step": 102870 + }, + { + "epoch": 29.9330811754437, + "grad_norm": 0.09440630674362183, + "learning_rate": 1.6388453378509383e-06, + "loss": 0.0103, + "step": 102880 + }, + { + "epoch": 29.935990689554846, + "grad_norm": 0.10546430200338364, + "learning_rate": 1.6369670650817004e-06, + "loss": 0.0085, + "step": 102890 + }, + { + "epoch": 29.938900203665987, + "grad_norm": 0.12097833305597305, + "learning_rate": 1.6350898071466225e-06, + "loss": 0.0092, + "step": 102900 + }, + { + "epoch": 29.941809717777133, + "grad_norm": 0.10115401446819305, + "learning_rate": 1.6332135641882757e-06, + "loss": 0.0113, + "step": 102910 + }, + { + "epoch": 29.944719231888275, + "grad_norm": 0.09432811290025711, + "learning_rate": 1.6313383363491424e-06, + "loss": 0.0121, + "step": 102920 + }, + { + "epoch": 29.947628745999417, + "grad_norm": 0.10614863783121109, + "learning_rate": 1.6294641237716373e-06, + "loss": 0.0114, + "step": 102930 + }, + { + "epoch": 29.950538260110562, + "grad_norm": 0.10308197885751724, + "learning_rate": 1.6275909265980982e-06, + "loss": 0.0096, + "step": 102940 + }, + { + "epoch": 29.953447774221704, + "grad_norm": 0.1290087252855301, + "learning_rate": 1.6257187449707745e-06, + "loss": 0.0091, + "step": 102950 + }, + { + "epoch": 29.95635728833285, + "grad_norm": 0.13921836018562317, + "learning_rate": 1.6238475790318492e-06, + "loss": 0.0119, + "step": 102960 + }, + { + "epoch": 29.95926680244399, + "grad_norm": 0.13044220209121704, + "learning_rate": 1.6219774289234285e-06, + "loss": 0.0112, + "step": 102970 + }, + { + "epoch": 29.962176316555134, + "grad_norm": 0.09974444657564163, + "learning_rate": 1.620108294787534e-06, + "loss": 0.0103, + "step": 102980 + }, + { + "epoch": 29.96508583066628, + "grad_norm": 0.12263932824134827, + "learning_rate": 1.618240176766112e-06, + "loss": 0.009, + "step": 102990 + }, + { + "epoch": 29.96799534477742, + "grad_norm": 0.09024327248334885, + "learning_rate": 1.6163730750010376e-06, + "loss": 0.0107, + "step": 103000 + }, + { + "epoch": 29.970904858888566, + "grad_norm": 0.11445499956607819, + "learning_rate": 1.6145069896341069e-06, + "loss": 0.0101, + "step": 103010 + }, + { + "epoch": 29.97381437299971, + "grad_norm": 0.1532578021287918, + "learning_rate": 1.6126419208070336e-06, + "loss": 0.0119, + "step": 103020 + }, + { + "epoch": 29.976723887110854, + "grad_norm": 0.12690888345241547, + "learning_rate": 1.610777868661451e-06, + "loss": 0.0134, + "step": 103030 + }, + { + "epoch": 29.979633401221996, + "grad_norm": 0.14670616388320923, + "learning_rate": 1.6089148333389342e-06, + "loss": 0.0097, + "step": 103040 + }, + { + "epoch": 29.982542915333138, + "grad_norm": 0.07866153120994568, + "learning_rate": 1.6070528149809632e-06, + "loss": 0.0103, + "step": 103050 + }, + { + "epoch": 29.985452429444283, + "grad_norm": 0.083701491355896, + "learning_rate": 1.6051918137289417e-06, + "loss": 0.011, + "step": 103060 + }, + { + "epoch": 29.988361943555425, + "grad_norm": 0.1177658662199974, + "learning_rate": 1.6033318297242033e-06, + "loss": 0.0139, + "step": 103070 + }, + { + "epoch": 29.99127145766657, + "grad_norm": 0.08951467275619507, + "learning_rate": 1.601472863108005e-06, + "loss": 0.0101, + "step": 103080 + }, + { + "epoch": 29.994180971777713, + "grad_norm": 0.0805337056517601, + "learning_rate": 1.5996149140215206e-06, + "loss": 0.0109, + "step": 103090 + }, + { + "epoch": 29.997090485888858, + "grad_norm": 0.09919805079698563, + "learning_rate": 1.597757982605847e-06, + "loss": 0.0106, + "step": 103100 + }, + { + "epoch": 30.0, + "grad_norm": 0.08726580440998077, + "learning_rate": 1.595902069002005e-06, + "loss": 0.0109, + "step": 103110 + }, + { + "epoch": 30.002909514111142, + "grad_norm": 0.08814594149589539, + "learning_rate": 1.5940471733509465e-06, + "loss": 0.0092, + "step": 103120 + }, + { + "epoch": 30.005819028222287, + "grad_norm": 0.12121989578008652, + "learning_rate": 1.5921932957935292e-06, + "loss": 0.0103, + "step": 103130 + }, + { + "epoch": 30.00872854233343, + "grad_norm": 0.07614349573850632, + "learning_rate": 1.5903404364705488e-06, + "loss": 0.0108, + "step": 103140 + }, + { + "epoch": 30.011638056444575, + "grad_norm": 0.07821208983659744, + "learning_rate": 1.5884885955227173e-06, + "loss": 0.011, + "step": 103150 + }, + { + "epoch": 30.014547570555717, + "grad_norm": 0.1182958260178566, + "learning_rate": 1.5866377730906678e-06, + "loss": 0.0111, + "step": 103160 + }, + { + "epoch": 30.017457084666862, + "grad_norm": 0.07547006011009216, + "learning_rate": 1.5847879693149609e-06, + "loss": 0.0097, + "step": 103170 + }, + { + "epoch": 30.020366598778004, + "grad_norm": 0.1237194836139679, + "learning_rate": 1.5829391843360707e-06, + "loss": 0.0102, + "step": 103180 + }, + { + "epoch": 30.023276112889146, + "grad_norm": 0.3577120304107666, + "learning_rate": 1.5810914182944085e-06, + "loss": 0.0118, + "step": 103190 + }, + { + "epoch": 30.02618562700029, + "grad_norm": 0.19005011022090912, + "learning_rate": 1.5792446713302937e-06, + "loss": 0.0106, + "step": 103200 + }, + { + "epoch": 30.029095141111434, + "grad_norm": 0.0836147665977478, + "learning_rate": 1.5773989435839792e-06, + "loss": 0.0098, + "step": 103210 + }, + { + "epoch": 30.03200465522258, + "grad_norm": 0.08408933877944946, + "learning_rate": 1.5755542351956308e-06, + "loss": 0.01, + "step": 103220 + }, + { + "epoch": 30.03491416933372, + "grad_norm": 0.09453455358743668, + "learning_rate": 1.5737105463053452e-06, + "loss": 0.0113, + "step": 103230 + }, + { + "epoch": 30.037823683444866, + "grad_norm": 0.0822778269648552, + "learning_rate": 1.5718678770531403e-06, + "loss": 0.0098, + "step": 103240 + }, + { + "epoch": 30.04073319755601, + "grad_norm": 0.1148940920829773, + "learning_rate": 1.5700262275789489e-06, + "loss": 0.0115, + "step": 103250 + }, + { + "epoch": 30.04364271166715, + "grad_norm": 0.10709510743618011, + "learning_rate": 1.5681855980226377e-06, + "loss": 0.0114, + "step": 103260 + }, + { + "epoch": 30.046552225778296, + "grad_norm": 0.13144202530384064, + "learning_rate": 1.5663459885239868e-06, + "loss": 0.0108, + "step": 103270 + }, + { + "epoch": 30.049461739889438, + "grad_norm": 0.09734860807657242, + "learning_rate": 1.5645073992227056e-06, + "loss": 0.0111, + "step": 103280 + }, + { + "epoch": 30.052371254000583, + "grad_norm": 0.09503549337387085, + "learning_rate": 1.5626698302584174e-06, + "loss": 0.0112, + "step": 103290 + }, + { + "epoch": 30.055280768111725, + "grad_norm": 0.11342082172632217, + "learning_rate": 1.5608332817706777e-06, + "loss": 0.0082, + "step": 103300 + }, + { + "epoch": 30.058190282222867, + "grad_norm": 0.10683822631835938, + "learning_rate": 1.558997753898963e-06, + "loss": 0.0114, + "step": 103310 + }, + { + "epoch": 30.061099796334013, + "grad_norm": 0.21133284270763397, + "learning_rate": 1.5571632467826652e-06, + "loss": 0.009, + "step": 103320 + }, + { + "epoch": 30.064009310445154, + "grad_norm": 0.18397383391857147, + "learning_rate": 1.5553297605610978e-06, + "loss": 0.0118, + "step": 103330 + }, + { + "epoch": 30.0669188245563, + "grad_norm": 0.11281268298625946, + "learning_rate": 1.5534972953735143e-06, + "loss": 0.0128, + "step": 103340 + }, + { + "epoch": 30.069828338667442, + "grad_norm": 0.12641794979572296, + "learning_rate": 1.551665851359072e-06, + "loss": 0.0118, + "step": 103350 + }, + { + "epoch": 30.072737852778587, + "grad_norm": 0.09797398746013641, + "learning_rate": 1.5498354286568545e-06, + "loss": 0.01, + "step": 103360 + }, + { + "epoch": 30.07564736688973, + "grad_norm": 0.12295550107955933, + "learning_rate": 1.5480060274058721e-06, + "loss": 0.0112, + "step": 103370 + }, + { + "epoch": 30.07855688100087, + "grad_norm": 0.09329306334257126, + "learning_rate": 1.546177647745059e-06, + "loss": 0.0099, + "step": 103380 + }, + { + "epoch": 30.081466395112017, + "grad_norm": 0.13914833962917328, + "learning_rate": 1.5443502898132672e-06, + "loss": 0.0094, + "step": 103390 + }, + { + "epoch": 30.08437590922316, + "grad_norm": 0.08172451704740524, + "learning_rate": 1.5425239537492674e-06, + "loss": 0.0106, + "step": 103400 + }, + { + "epoch": 30.087285423334304, + "grad_norm": 0.17728687822818756, + "learning_rate": 1.5406986396917622e-06, + "loss": 0.0111, + "step": 103410 + }, + { + "epoch": 30.090194937445446, + "grad_norm": 0.15300533175468445, + "learning_rate": 1.5388743477793737e-06, + "loss": 0.0112, + "step": 103420 + }, + { + "epoch": 30.09310445155659, + "grad_norm": 0.1140894889831543, + "learning_rate": 1.5370510781506415e-06, + "loss": 0.0101, + "step": 103430 + }, + { + "epoch": 30.096013965667733, + "grad_norm": 0.14600563049316406, + "learning_rate": 1.535228830944031e-06, + "loss": 0.0103, + "step": 103440 + }, + { + "epoch": 30.098923479778875, + "grad_norm": 0.09308206290006638, + "learning_rate": 1.5334076062979335e-06, + "loss": 0.0111, + "step": 103450 + }, + { + "epoch": 30.10183299389002, + "grad_norm": 0.1037123054265976, + "learning_rate": 1.5315874043506584e-06, + "loss": 0.0115, + "step": 103460 + }, + { + "epoch": 30.104742508001163, + "grad_norm": 0.09141115099191666, + "learning_rate": 1.5297682252404316e-06, + "loss": 0.0113, + "step": 103470 + }, + { + "epoch": 30.10765202211231, + "grad_norm": 0.14962445199489594, + "learning_rate": 1.5279500691054127e-06, + "loss": 0.0099, + "step": 103480 + }, + { + "epoch": 30.11056153622345, + "grad_norm": 0.08548431098461151, + "learning_rate": 1.5261329360836813e-06, + "loss": 0.0094, + "step": 103490 + }, + { + "epoch": 30.113471050334596, + "grad_norm": 0.10140138864517212, + "learning_rate": 1.52431682631323e-06, + "loss": 0.012, + "step": 103500 + }, + { + "epoch": 30.116380564445738, + "grad_norm": 0.1146552711725235, + "learning_rate": 1.522501739931989e-06, + "loss": 0.0109, + "step": 103510 + }, + { + "epoch": 30.11929007855688, + "grad_norm": 0.10760214924812317, + "learning_rate": 1.5206876770777928e-06, + "loss": 0.0098, + "step": 103520 + }, + { + "epoch": 30.122199592668025, + "grad_norm": 0.07790767401456833, + "learning_rate": 1.5188746378884111e-06, + "loss": 0.0104, + "step": 103530 + }, + { + "epoch": 30.125109106779167, + "grad_norm": 0.10486675798892975, + "learning_rate": 1.517062622501537e-06, + "loss": 0.0107, + "step": 103540 + }, + { + "epoch": 30.128018620890312, + "grad_norm": 0.11626142263412476, + "learning_rate": 1.5152516310547727e-06, + "loss": 0.0105, + "step": 103550 + }, + { + "epoch": 30.130928135001454, + "grad_norm": 0.12500493228435516, + "learning_rate": 1.5134416636856607e-06, + "loss": 0.0105, + "step": 103560 + }, + { + "epoch": 30.1338376491126, + "grad_norm": 0.05703049153089523, + "learning_rate": 1.5116327205316465e-06, + "loss": 0.0098, + "step": 103570 + }, + { + "epoch": 30.136747163223742, + "grad_norm": 0.09867025911808014, + "learning_rate": 1.5098248017301136e-06, + "loss": 0.0118, + "step": 103580 + }, + { + "epoch": 30.139656677334884, + "grad_norm": 0.14735692739486694, + "learning_rate": 1.5080179074183587e-06, + "loss": 0.0106, + "step": 103590 + }, + { + "epoch": 30.14256619144603, + "grad_norm": 0.07283208519220352, + "learning_rate": 1.5062120377336036e-06, + "loss": 0.0094, + "step": 103600 + }, + { + "epoch": 30.14547570555717, + "grad_norm": 0.06981806457042694, + "learning_rate": 1.5044071928129953e-06, + "loss": 0.011, + "step": 103610 + }, + { + "epoch": 30.148385219668317, + "grad_norm": 0.13346993923187256, + "learning_rate": 1.5026033727935978e-06, + "loss": 0.0094, + "step": 103620 + }, + { + "epoch": 30.15129473377946, + "grad_norm": 0.09835748374462128, + "learning_rate": 1.5008005778123963e-06, + "loss": 0.012, + "step": 103630 + }, + { + "epoch": 30.154204247890604, + "grad_norm": 0.13341672718524933, + "learning_rate": 1.4989988080063043e-06, + "loss": 0.0111, + "step": 103640 + }, + { + "epoch": 30.157113762001746, + "grad_norm": 0.1406274288892746, + "learning_rate": 1.4971980635121562e-06, + "loss": 0.0115, + "step": 103650 + }, + { + "epoch": 30.160023276112888, + "grad_norm": 0.11412177234888077, + "learning_rate": 1.4953983444667025e-06, + "loss": 0.0078, + "step": 103660 + }, + { + "epoch": 30.162932790224033, + "grad_norm": 0.12772245705127716, + "learning_rate": 1.4935996510066203e-06, + "loss": 0.0132, + "step": 103670 + }, + { + "epoch": 30.165842304335175, + "grad_norm": 0.11617162823677063, + "learning_rate": 1.4918019832685136e-06, + "loss": 0.0135, + "step": 103680 + }, + { + "epoch": 30.16875181844632, + "grad_norm": 0.13743850588798523, + "learning_rate": 1.490005341388902e-06, + "loss": 0.0108, + "step": 103690 + }, + { + "epoch": 30.171661332557463, + "grad_norm": 0.0694182813167572, + "learning_rate": 1.488209725504221e-06, + "loss": 0.0098, + "step": 103700 + }, + { + "epoch": 30.174570846668605, + "grad_norm": 0.08494656533002853, + "learning_rate": 1.4864151357508433e-06, + "loss": 0.0106, + "step": 103710 + }, + { + "epoch": 30.17748036077975, + "grad_norm": 0.15668514370918274, + "learning_rate": 1.4846215722650547e-06, + "loss": 0.0119, + "step": 103720 + }, + { + "epoch": 30.180389874890892, + "grad_norm": 0.08049424737691879, + "learning_rate": 1.4828290351830665e-06, + "loss": 0.0099, + "step": 103730 + }, + { + "epoch": 30.183299389002038, + "grad_norm": 0.12102431803941727, + "learning_rate": 1.4810375246409996e-06, + "loss": 0.0117, + "step": 103740 + }, + { + "epoch": 30.18620890311318, + "grad_norm": 0.09686911851167679, + "learning_rate": 1.479247040774922e-06, + "loss": 0.0099, + "step": 103750 + }, + { + "epoch": 30.189118417224325, + "grad_norm": 0.07911285758018494, + "learning_rate": 1.4774575837208032e-06, + "loss": 0.0106, + "step": 103760 + }, + { + "epoch": 30.192027931335467, + "grad_norm": 0.24151480197906494, + "learning_rate": 1.475669153614536e-06, + "loss": 0.0105, + "step": 103770 + }, + { + "epoch": 30.19493744544661, + "grad_norm": 0.09205321967601776, + "learning_rate": 1.473881750591944e-06, + "loss": 0.0092, + "step": 103780 + }, + { + "epoch": 30.197846959557754, + "grad_norm": 0.11365154385566711, + "learning_rate": 1.472095374788771e-06, + "loss": 0.011, + "step": 103790 + }, + { + "epoch": 30.200756473668896, + "grad_norm": 0.11532478034496307, + "learning_rate": 1.470310026340676e-06, + "loss": 0.0126, + "step": 103800 + }, + { + "epoch": 30.20366598778004, + "grad_norm": 0.09694591909646988, + "learning_rate": 1.4685257053832485e-06, + "loss": 0.0102, + "step": 103810 + }, + { + "epoch": 30.206575501891184, + "grad_norm": 0.08770394325256348, + "learning_rate": 1.46674241205199e-06, + "loss": 0.01, + "step": 103820 + }, + { + "epoch": 30.20948501600233, + "grad_norm": 0.2188090980052948, + "learning_rate": 1.4649601464823376e-06, + "loss": 0.0102, + "step": 103830 + }, + { + "epoch": 30.21239453011347, + "grad_norm": 0.09261366724967957, + "learning_rate": 1.463178908809636e-06, + "loss": 0.0112, + "step": 103840 + }, + { + "epoch": 30.215304044224613, + "grad_norm": 0.08838114887475967, + "learning_rate": 1.4613986991691603e-06, + "loss": 0.0108, + "step": 103850 + }, + { + "epoch": 30.21821355833576, + "grad_norm": 0.11517535895109177, + "learning_rate": 1.4596195176961092e-06, + "loss": 0.0131, + "step": 103860 + }, + { + "epoch": 30.2211230724469, + "grad_norm": 0.09030836820602417, + "learning_rate": 1.4578413645255945e-06, + "loss": 0.0098, + "step": 103870 + }, + { + "epoch": 30.224032586558046, + "grad_norm": 0.08262136578559875, + "learning_rate": 1.4560642397926599e-06, + "loss": 0.0103, + "step": 103880 + }, + { + "epoch": 30.226942100669188, + "grad_norm": 0.11506786942481995, + "learning_rate": 1.4542881436322625e-06, + "loss": 0.0111, + "step": 103890 + }, + { + "epoch": 30.229851614780333, + "grad_norm": 0.09458038955926895, + "learning_rate": 1.452513076179286e-06, + "loss": 0.0096, + "step": 103900 + }, + { + "epoch": 30.232761128891475, + "grad_norm": 0.08717934042215347, + "learning_rate": 1.4507390375685376e-06, + "loss": 0.0096, + "step": 103910 + }, + { + "epoch": 30.235670643002617, + "grad_norm": 0.12714262306690216, + "learning_rate": 1.4489660279347427e-06, + "loss": 0.0106, + "step": 103920 + }, + { + "epoch": 30.238580157113763, + "grad_norm": 0.08668507635593414, + "learning_rate": 1.4471940474125472e-06, + "loss": 0.0111, + "step": 103930 + }, + { + "epoch": 30.241489671224905, + "grad_norm": 0.09792734682559967, + "learning_rate": 1.4454230961365217e-06, + "loss": 0.0122, + "step": 103940 + }, + { + "epoch": 30.24439918533605, + "grad_norm": 0.18155668675899506, + "learning_rate": 1.4436531742411617e-06, + "loss": 0.0129, + "step": 103950 + }, + { + "epoch": 30.247308699447192, + "grad_norm": 0.11382405459880829, + "learning_rate": 1.4418842818608785e-06, + "loss": 0.0107, + "step": 103960 + }, + { + "epoch": 30.250218213558334, + "grad_norm": 0.08610507100820541, + "learning_rate": 1.4401164191300076e-06, + "loss": 0.0103, + "step": 103970 + }, + { + "epoch": 30.25312772766948, + "grad_norm": 0.127118319272995, + "learning_rate": 1.4383495861828083e-06, + "loss": 0.0104, + "step": 103980 + }, + { + "epoch": 30.25603724178062, + "grad_norm": 0.06659460067749023, + "learning_rate": 1.43658378315346e-06, + "loss": 0.0094, + "step": 103990 + }, + { + "epoch": 30.258946755891767, + "grad_norm": 0.13005906343460083, + "learning_rate": 1.4348190101760606e-06, + "loss": 0.0098, + "step": 104000 + }, + { + "epoch": 30.26185627000291, + "grad_norm": 0.0763126015663147, + "learning_rate": 1.4330552673846342e-06, + "loss": 0.0087, + "step": 104010 + }, + { + "epoch": 30.264765784114054, + "grad_norm": 0.11550036072731018, + "learning_rate": 1.4312925549131289e-06, + "loss": 0.0105, + "step": 104020 + }, + { + "epoch": 30.267675298225196, + "grad_norm": 0.1373172104358673, + "learning_rate": 1.4295308728954093e-06, + "loss": 0.011, + "step": 104030 + }, + { + "epoch": 30.270584812336338, + "grad_norm": 0.09596370160579681, + "learning_rate": 1.4277702214652566e-06, + "loss": 0.0104, + "step": 104040 + }, + { + "epoch": 30.273494326447484, + "grad_norm": 0.08254432678222656, + "learning_rate": 1.4260106007563906e-06, + "loss": 0.0093, + "step": 104050 + }, + { + "epoch": 30.276403840558626, + "grad_norm": 0.11744855344295502, + "learning_rate": 1.424252010902441e-06, + "loss": 0.0097, + "step": 104060 + }, + { + "epoch": 30.27931335466977, + "grad_norm": 0.11121942102909088, + "learning_rate": 1.422494452036956e-06, + "loss": 0.009, + "step": 104070 + }, + { + "epoch": 30.282222868780913, + "grad_norm": 0.07434450089931488, + "learning_rate": 1.4207379242934121e-06, + "loss": 0.0103, + "step": 104080 + }, + { + "epoch": 30.28513238289206, + "grad_norm": 0.11534038931131363, + "learning_rate": 1.418982427805211e-06, + "loss": 0.0101, + "step": 104090 + }, + { + "epoch": 30.2880418970032, + "grad_norm": 0.10008417814970016, + "learning_rate": 1.4172279627056695e-06, + "loss": 0.0102, + "step": 104100 + }, + { + "epoch": 30.290951411114342, + "grad_norm": 0.11260826885700226, + "learning_rate": 1.4154745291280209e-06, + "loss": 0.0106, + "step": 104110 + }, + { + "epoch": 30.293860925225488, + "grad_norm": 0.10742580890655518, + "learning_rate": 1.4137221272054318e-06, + "loss": 0.0103, + "step": 104120 + }, + { + "epoch": 30.29677043933663, + "grad_norm": 0.0617227703332901, + "learning_rate": 1.4119707570709895e-06, + "loss": 0.0105, + "step": 104130 + }, + { + "epoch": 30.299679953447775, + "grad_norm": 0.06543669104576111, + "learning_rate": 1.4102204188576905e-06, + "loss": 0.0125, + "step": 104140 + }, + { + "epoch": 30.302589467558917, + "grad_norm": 0.13847072422504425, + "learning_rate": 1.4084711126984674e-06, + "loss": 0.0105, + "step": 104150 + }, + { + "epoch": 30.305498981670063, + "grad_norm": 0.09959860146045685, + "learning_rate": 1.4067228387261682e-06, + "loss": 0.0106, + "step": 104160 + }, + { + "epoch": 30.308408495781205, + "grad_norm": 0.09151283651590347, + "learning_rate": 1.4049755970735606e-06, + "loss": 0.0097, + "step": 104170 + }, + { + "epoch": 30.311318009892346, + "grad_norm": 0.0935925841331482, + "learning_rate": 1.4032293878733398e-06, + "loss": 0.0112, + "step": 104180 + }, + { + "epoch": 30.314227524003492, + "grad_norm": 0.10972609370946884, + "learning_rate": 1.4014842112581117e-06, + "loss": 0.0105, + "step": 104190 + }, + { + "epoch": 30.317137038114634, + "grad_norm": 0.11533210426568985, + "learning_rate": 1.39974006736042e-06, + "loss": 0.0102, + "step": 104200 + }, + { + "epoch": 30.32004655222578, + "grad_norm": 0.15704567730426788, + "learning_rate": 1.397996956312712e-06, + "loss": 0.0119, + "step": 104210 + }, + { + "epoch": 30.32295606633692, + "grad_norm": 0.10478635877370834, + "learning_rate": 1.3962548782473723e-06, + "loss": 0.0097, + "step": 104220 + }, + { + "epoch": 30.325865580448067, + "grad_norm": 0.10519915819168091, + "learning_rate": 1.3945138332966962e-06, + "loss": 0.0103, + "step": 104230 + }, + { + "epoch": 30.32877509455921, + "grad_norm": 0.05585957318544388, + "learning_rate": 1.3927738215929064e-06, + "loss": 0.0094, + "step": 104240 + }, + { + "epoch": 30.33168460867035, + "grad_norm": 0.1267285794019699, + "learning_rate": 1.3910348432681458e-06, + "loss": 0.0101, + "step": 104250 + }, + { + "epoch": 30.334594122781496, + "grad_norm": 0.08199922740459442, + "learning_rate": 1.3892968984544767e-06, + "loss": 0.0104, + "step": 104260 + }, + { + "epoch": 30.337503636892638, + "grad_norm": 0.22397685050964355, + "learning_rate": 1.387559987283884e-06, + "loss": 0.0115, + "step": 104270 + }, + { + "epoch": 30.340413151003784, + "grad_norm": 0.07445159554481506, + "learning_rate": 1.38582410988828e-06, + "loss": 0.0097, + "step": 104280 + }, + { + "epoch": 30.343322665114925, + "grad_norm": 0.0937233492732048, + "learning_rate": 1.384089266399488e-06, + "loss": 0.0106, + "step": 104290 + }, + { + "epoch": 30.34623217922607, + "grad_norm": 0.10164239257574081, + "learning_rate": 1.3823554569492557e-06, + "loss": 0.009, + "step": 104300 + }, + { + "epoch": 30.349141693337213, + "grad_norm": 0.1221945658326149, + "learning_rate": 1.3806226816692595e-06, + "loss": 0.011, + "step": 104310 + }, + { + "epoch": 30.352051207448355, + "grad_norm": 0.15211069583892822, + "learning_rate": 1.3788909406910927e-06, + "loss": 0.0113, + "step": 104320 + }, + { + "epoch": 30.3549607215595, + "grad_norm": 0.12965542078018188, + "learning_rate": 1.3771602341462664e-06, + "loss": 0.0115, + "step": 104330 + }, + { + "epoch": 30.357870235670642, + "grad_norm": 0.11057061702013016, + "learning_rate": 1.3754305621662138e-06, + "loss": 0.0109, + "step": 104340 + }, + { + "epoch": 30.360779749781788, + "grad_norm": 0.12737445533275604, + "learning_rate": 1.3737019248823e-06, + "loss": 0.0108, + "step": 104350 + }, + { + "epoch": 30.36368926389293, + "grad_norm": 0.07845453917980194, + "learning_rate": 1.3719743224257986e-06, + "loss": 0.0097, + "step": 104360 + }, + { + "epoch": 30.36659877800407, + "grad_norm": 0.09377983212471008, + "learning_rate": 1.3702477549279092e-06, + "loss": 0.0097, + "step": 104370 + }, + { + "epoch": 30.369508292115217, + "grad_norm": 0.13006684184074402, + "learning_rate": 1.3685222225197536e-06, + "loss": 0.0112, + "step": 104380 + }, + { + "epoch": 30.37241780622636, + "grad_norm": 0.127424955368042, + "learning_rate": 1.366797725332377e-06, + "loss": 0.0109, + "step": 104390 + }, + { + "epoch": 30.375327320337504, + "grad_norm": 0.07106469571590424, + "learning_rate": 1.3650742634967433e-06, + "loss": 0.0104, + "step": 104400 + }, + { + "epoch": 30.378236834448646, + "grad_norm": 0.15895120799541473, + "learning_rate": 1.3633518371437326e-06, + "loss": 0.0125, + "step": 104410 + }, + { + "epoch": 30.381146348559792, + "grad_norm": 0.05815383046865463, + "learning_rate": 1.3616304464041568e-06, + "loss": 0.0104, + "step": 104420 + }, + { + "epoch": 30.384055862670934, + "grad_norm": 0.12298205494880676, + "learning_rate": 1.3599100914087465e-06, + "loss": 0.0094, + "step": 104430 + }, + { + "epoch": 30.386965376782076, + "grad_norm": 0.19083473086357117, + "learning_rate": 1.3581907722881436e-06, + "loss": 0.0108, + "step": 104440 + }, + { + "epoch": 30.38987489089322, + "grad_norm": 0.1595262885093689, + "learning_rate": 1.356472489172924e-06, + "loss": 0.0105, + "step": 104450 + }, + { + "epoch": 30.392784405004363, + "grad_norm": 0.15094433724880219, + "learning_rate": 1.3547552421935832e-06, + "loss": 0.0113, + "step": 104460 + }, + { + "epoch": 30.39569391911551, + "grad_norm": 0.213905930519104, + "learning_rate": 1.3530390314805303e-06, + "loss": 0.0104, + "step": 104470 + }, + { + "epoch": 30.39860343322665, + "grad_norm": 0.08591470122337341, + "learning_rate": 1.3513238571640973e-06, + "loss": 0.0094, + "step": 104480 + }, + { + "epoch": 30.401512947337796, + "grad_norm": 0.09995967149734497, + "learning_rate": 1.3496097193745437e-06, + "loss": 0.0097, + "step": 104490 + }, + { + "epoch": 30.404422461448938, + "grad_norm": 0.11659444868564606, + "learning_rate": 1.3478966182420522e-06, + "loss": 0.0115, + "step": 104500 + }, + { + "epoch": 30.40733197556008, + "grad_norm": 0.10706035047769547, + "learning_rate": 1.3461845538967115e-06, + "loss": 0.0103, + "step": 104510 + }, + { + "epoch": 30.410241489671225, + "grad_norm": 0.13555824756622314, + "learning_rate": 1.3444735264685497e-06, + "loss": 0.0104, + "step": 104520 + }, + { + "epoch": 30.413151003782367, + "grad_norm": 0.11128878593444824, + "learning_rate": 1.342763536087503e-06, + "loss": 0.01, + "step": 104530 + }, + { + "epoch": 30.416060517893513, + "grad_norm": 0.13133321702480316, + "learning_rate": 1.3410545828834352e-06, + "loss": 0.0095, + "step": 104540 + }, + { + "epoch": 30.418970032004655, + "grad_norm": 0.10979580879211426, + "learning_rate": 1.339346666986135e-06, + "loss": 0.0107, + "step": 104550 + }, + { + "epoch": 30.4218795461158, + "grad_norm": 0.13786691427230835, + "learning_rate": 1.3376397885252977e-06, + "loss": 0.0119, + "step": 104560 + }, + { + "epoch": 30.424789060226942, + "grad_norm": 0.1364724040031433, + "learning_rate": 1.3359339476305589e-06, + "loss": 0.0122, + "step": 104570 + }, + { + "epoch": 30.427698574338084, + "grad_norm": 0.14525219798088074, + "learning_rate": 1.3342291444314592e-06, + "loss": 0.0098, + "step": 104580 + }, + { + "epoch": 30.43060808844923, + "grad_norm": 0.09344445914030075, + "learning_rate": 1.332525379057472e-06, + "loss": 0.0098, + "step": 104590 + }, + { + "epoch": 30.43351760256037, + "grad_norm": 0.133214071393013, + "learning_rate": 1.3308226516379824e-06, + "loss": 0.0111, + "step": 104600 + }, + { + "epoch": 30.436427116671517, + "grad_norm": 0.1435423046350479, + "learning_rate": 1.3291209623023032e-06, + "loss": 0.0098, + "step": 104610 + }, + { + "epoch": 30.43933663078266, + "grad_norm": 0.10436933487653732, + "learning_rate": 1.3274203111796695e-06, + "loss": 0.0078, + "step": 104620 + }, + { + "epoch": 30.442246144893804, + "grad_norm": 0.11242033541202545, + "learning_rate": 1.325720698399233e-06, + "loss": 0.01, + "step": 104630 + }, + { + "epoch": 30.445155659004946, + "grad_norm": 0.11749500036239624, + "learning_rate": 1.3240221240900607e-06, + "loss": 0.0093, + "step": 104640 + }, + { + "epoch": 30.44806517311609, + "grad_norm": 0.11556417495012283, + "learning_rate": 1.3223245883811592e-06, + "loss": 0.0132, + "step": 104650 + }, + { + "epoch": 30.450974687227234, + "grad_norm": 0.08100299537181854, + "learning_rate": 1.3206280914014418e-06, + "loss": 0.0093, + "step": 104660 + }, + { + "epoch": 30.453884201338376, + "grad_norm": 0.10251624137163162, + "learning_rate": 1.3189326332797425e-06, + "loss": 0.0106, + "step": 104670 + }, + { + "epoch": 30.45679371544952, + "grad_norm": 0.10258842259645462, + "learning_rate": 1.3172382141448214e-06, + "loss": 0.0107, + "step": 104680 + }, + { + "epoch": 30.459703229560663, + "grad_norm": 0.08390691876411438, + "learning_rate": 1.3155448341253623e-06, + "loss": 0.0111, + "step": 104690 + }, + { + "epoch": 30.462612743671805, + "grad_norm": 0.07769119739532471, + "learning_rate": 1.3138524933499641e-06, + "loss": 0.0114, + "step": 104700 + }, + { + "epoch": 30.46552225778295, + "grad_norm": 0.11187168955802917, + "learning_rate": 1.3121611919471455e-06, + "loss": 0.0119, + "step": 104710 + }, + { + "epoch": 30.468431771894092, + "grad_norm": 0.1140156090259552, + "learning_rate": 1.310470930045352e-06, + "loss": 0.0126, + "step": 104720 + }, + { + "epoch": 30.471341286005238, + "grad_norm": 0.08819785714149475, + "learning_rate": 1.3087817077729513e-06, + "loss": 0.0128, + "step": 104730 + }, + { + "epoch": 30.47425080011638, + "grad_norm": 0.10603176802396774, + "learning_rate": 1.307093525258224e-06, + "loss": 0.0122, + "step": 104740 + }, + { + "epoch": 30.477160314227525, + "grad_norm": 0.08151881396770477, + "learning_rate": 1.3054063826293771e-06, + "loss": 0.0089, + "step": 104750 + }, + { + "epoch": 30.480069828338667, + "grad_norm": 0.14047616720199585, + "learning_rate": 1.303720280014542e-06, + "loss": 0.0122, + "step": 104760 + }, + { + "epoch": 30.48297934244981, + "grad_norm": 0.10781526565551758, + "learning_rate": 1.3020352175417626e-06, + "loss": 0.0101, + "step": 104770 + }, + { + "epoch": 30.485888856560955, + "grad_norm": 0.1114276573061943, + "learning_rate": 1.3003511953390084e-06, + "loss": 0.0085, + "step": 104780 + }, + { + "epoch": 30.488798370672097, + "grad_norm": 0.12103437632322311, + "learning_rate": 1.2986682135341722e-06, + "loss": 0.0107, + "step": 104790 + }, + { + "epoch": 30.491707884783242, + "grad_norm": 0.09844201803207397, + "learning_rate": 1.2969862722550647e-06, + "loss": 0.0101, + "step": 104800 + }, + { + "epoch": 30.494617398894384, + "grad_norm": 0.1580093502998352, + "learning_rate": 1.2953053716294173e-06, + "loss": 0.0103, + "step": 104810 + }, + { + "epoch": 30.49752691300553, + "grad_norm": 0.12066905200481415, + "learning_rate": 1.2936255117848843e-06, + "loss": 0.0101, + "step": 104820 + }, + { + "epoch": 30.50043642711667, + "grad_norm": 0.10575578361749649, + "learning_rate": 1.2919466928490392e-06, + "loss": 0.0101, + "step": 104830 + }, + { + "epoch": 30.503345941227813, + "grad_norm": 0.08909182250499725, + "learning_rate": 1.2902689149493779e-06, + "loss": 0.0114, + "step": 104840 + }, + { + "epoch": 30.50625545533896, + "grad_norm": 0.11060108989477158, + "learning_rate": 1.2885921782133153e-06, + "loss": 0.0112, + "step": 104850 + }, + { + "epoch": 30.5091649694501, + "grad_norm": 0.1166629046201706, + "learning_rate": 1.286916482768188e-06, + "loss": 0.0108, + "step": 104860 + }, + { + "epoch": 30.512074483561246, + "grad_norm": 0.07995465397834778, + "learning_rate": 1.2852418287412588e-06, + "loss": 0.0097, + "step": 104870 + }, + { + "epoch": 30.51498399767239, + "grad_norm": 0.10147503018379211, + "learning_rate": 1.2835682162596997e-06, + "loss": 0.013, + "step": 104880 + }, + { + "epoch": 30.517893511783534, + "grad_norm": 0.13056348264217377, + "learning_rate": 1.281895645450617e-06, + "loss": 0.0097, + "step": 104890 + }, + { + "epoch": 30.520803025894676, + "grad_norm": 0.1107187420129776, + "learning_rate": 1.2802241164410262e-06, + "loss": 0.0095, + "step": 104900 + }, + { + "epoch": 30.523712540005818, + "grad_norm": 0.15526622533798218, + "learning_rate": 1.2785536293578703e-06, + "loss": 0.0105, + "step": 104910 + }, + { + "epoch": 30.526622054116963, + "grad_norm": 0.07806511968374252, + "learning_rate": 1.2768841843280165e-06, + "loss": 0.0104, + "step": 104920 + }, + { + "epoch": 30.529531568228105, + "grad_norm": 0.2299979031085968, + "learning_rate": 1.2752157814782433e-06, + "loss": 0.011, + "step": 104930 + }, + { + "epoch": 30.53244108233925, + "grad_norm": 0.09685398638248444, + "learning_rate": 1.2735484209352528e-06, + "loss": 0.0115, + "step": 104940 + }, + { + "epoch": 30.535350596450392, + "grad_norm": 0.10169509798288345, + "learning_rate": 1.271882102825675e-06, + "loss": 0.0095, + "step": 104950 + }, + { + "epoch": 30.538260110561538, + "grad_norm": 0.08994588255882263, + "learning_rate": 1.2702168272760538e-06, + "loss": 0.0108, + "step": 104960 + }, + { + "epoch": 30.54116962467268, + "grad_norm": 0.38209179043769836, + "learning_rate": 1.268552594412855e-06, + "loss": 0.0103, + "step": 104970 + }, + { + "epoch": 30.54407913878382, + "grad_norm": 0.152121901512146, + "learning_rate": 1.2668894043624657e-06, + "loss": 0.0117, + "step": 104980 + }, + { + "epoch": 30.546988652894967, + "grad_norm": 0.09875912219285965, + "learning_rate": 1.2652272572511998e-06, + "loss": 0.0131, + "step": 104990 + }, + { + "epoch": 30.54989816700611, + "grad_norm": 0.1176685020327568, + "learning_rate": 1.2635661532052812e-06, + "loss": 0.0115, + "step": 105000 + }, + { + "epoch": 30.552807681117255, + "grad_norm": 0.1314648687839508, + "learning_rate": 1.2619060923508574e-06, + "loss": 0.0098, + "step": 105010 + }, + { + "epoch": 30.555717195228397, + "grad_norm": 0.22043214738368988, + "learning_rate": 1.2602470748140028e-06, + "loss": 0.0097, + "step": 105020 + }, + { + "epoch": 30.558626709339542, + "grad_norm": 0.13740983605384827, + "learning_rate": 1.2585891007207113e-06, + "loss": 0.0098, + "step": 105030 + }, + { + "epoch": 30.561536223450684, + "grad_norm": 0.08657616376876831, + "learning_rate": 1.2569321701968922e-06, + "loss": 0.0098, + "step": 105040 + }, + { + "epoch": 30.564445737561826, + "grad_norm": 0.11109429597854614, + "learning_rate": 1.2552762833683735e-06, + "loss": 0.0106, + "step": 105050 + }, + { + "epoch": 30.56735525167297, + "grad_norm": 0.07598654180765152, + "learning_rate": 1.2536214403609193e-06, + "loss": 0.0088, + "step": 105060 + }, + { + "epoch": 30.570264765784113, + "grad_norm": 0.0785161480307579, + "learning_rate": 1.2519676413001974e-06, + "loss": 0.0103, + "step": 105070 + }, + { + "epoch": 30.57317427989526, + "grad_norm": 0.1078147143125534, + "learning_rate": 1.2503148863118008e-06, + "loss": 0.01, + "step": 105080 + }, + { + "epoch": 30.5760837940064, + "grad_norm": 0.08934729546308517, + "learning_rate": 1.2486631755212507e-06, + "loss": 0.0103, + "step": 105090 + }, + { + "epoch": 30.578993308117543, + "grad_norm": 0.09432261437177658, + "learning_rate": 1.2470125090539819e-06, + "loss": 0.01, + "step": 105100 + }, + { + "epoch": 30.581902822228688, + "grad_norm": 0.08351969718933105, + "learning_rate": 1.2453628870353506e-06, + "loss": 0.0095, + "step": 105110 + }, + { + "epoch": 30.58481233633983, + "grad_norm": 0.129583477973938, + "learning_rate": 1.2437143095906367e-06, + "loss": 0.0138, + "step": 105120 + }, + { + "epoch": 30.587721850450976, + "grad_norm": 0.08456286042928696, + "learning_rate": 1.242066776845035e-06, + "loss": 0.0095, + "step": 105130 + }, + { + "epoch": 30.590631364562118, + "grad_norm": 0.10005814582109451, + "learning_rate": 1.2404202889236705e-06, + "loss": 0.0113, + "step": 105140 + }, + { + "epoch": 30.593540878673263, + "grad_norm": 0.11066337674856186, + "learning_rate": 1.2387748459515763e-06, + "loss": 0.0089, + "step": 105150 + }, + { + "epoch": 30.596450392784405, + "grad_norm": 0.2754073143005371, + "learning_rate": 1.237130448053716e-06, + "loss": 0.0097, + "step": 105160 + }, + { + "epoch": 30.599359906895547, + "grad_norm": 0.09211724996566772, + "learning_rate": 1.2354870953549745e-06, + "loss": 0.0096, + "step": 105170 + }, + { + "epoch": 30.602269421006692, + "grad_norm": 0.1400904506444931, + "learning_rate": 1.2338447879801472e-06, + "loss": 0.0116, + "step": 105180 + }, + { + "epoch": 30.605178935117834, + "grad_norm": 0.12306106090545654, + "learning_rate": 1.2322035260539627e-06, + "loss": 0.012, + "step": 105190 + }, + { + "epoch": 30.60808844922898, + "grad_norm": 0.06914076954126358, + "learning_rate": 1.2305633097010577e-06, + "loss": 0.0106, + "step": 105200 + }, + { + "epoch": 30.61099796334012, + "grad_norm": 0.12039081007242203, + "learning_rate": 1.2289241390459994e-06, + "loss": 0.0112, + "step": 105210 + }, + { + "epoch": 30.613907477451267, + "grad_norm": 0.0933828130364418, + "learning_rate": 1.2272860142132731e-06, + "loss": 0.0102, + "step": 105220 + }, + { + "epoch": 30.61681699156241, + "grad_norm": 0.1234767884016037, + "learning_rate": 1.2256489353272843e-06, + "loss": 0.0099, + "step": 105230 + }, + { + "epoch": 30.61972650567355, + "grad_norm": 0.10486501455307007, + "learning_rate": 1.224012902512352e-06, + "loss": 0.0108, + "step": 105240 + }, + { + "epoch": 30.622636019784697, + "grad_norm": 0.11062643676996231, + "learning_rate": 1.222377915892725e-06, + "loss": 0.0088, + "step": 105250 + }, + { + "epoch": 30.62554553389584, + "grad_norm": 0.09878414124250412, + "learning_rate": 1.2207439755925743e-06, + "loss": 0.0104, + "step": 105260 + }, + { + "epoch": 30.628455048006984, + "grad_norm": 0.08611579984426498, + "learning_rate": 1.2191110817359818e-06, + "loss": 0.0112, + "step": 105270 + }, + { + "epoch": 30.631364562118126, + "grad_norm": 0.12052283436059952, + "learning_rate": 1.2174792344469554e-06, + "loss": 0.0103, + "step": 105280 + }, + { + "epoch": 30.63427407622927, + "grad_norm": 0.08477263897657394, + "learning_rate": 1.215848433849429e-06, + "loss": 0.0087, + "step": 105290 + }, + { + "epoch": 30.637183590340413, + "grad_norm": 0.10584204643964767, + "learning_rate": 1.214218680067244e-06, + "loss": 0.0112, + "step": 105300 + }, + { + "epoch": 30.640093104451555, + "grad_norm": 0.17931678891181946, + "learning_rate": 1.2125899732241709e-06, + "loss": 0.0115, + "step": 105310 + }, + { + "epoch": 30.6430026185627, + "grad_norm": 0.09551700949668884, + "learning_rate": 1.2109623134439007e-06, + "loss": 0.0103, + "step": 105320 + }, + { + "epoch": 30.645912132673843, + "grad_norm": 0.1928028166294098, + "learning_rate": 1.209335700850045e-06, + "loss": 0.0112, + "step": 105330 + }, + { + "epoch": 30.648821646784988, + "grad_norm": 0.1966690868139267, + "learning_rate": 1.207710135566133e-06, + "loss": 0.0117, + "step": 105340 + }, + { + "epoch": 30.65173116089613, + "grad_norm": 0.09361927211284637, + "learning_rate": 1.2060856177156092e-06, + "loss": 0.0092, + "step": 105350 + }, + { + "epoch": 30.654640675007272, + "grad_norm": 0.09442639350891113, + "learning_rate": 1.2044621474218553e-06, + "loss": 0.0115, + "step": 105360 + }, + { + "epoch": 30.657550189118417, + "grad_norm": 0.1732500046491623, + "learning_rate": 1.202839724808159e-06, + "loss": 0.0088, + "step": 105370 + }, + { + "epoch": 30.66045970322956, + "grad_norm": 0.11532370001077652, + "learning_rate": 1.2012183499977303e-06, + "loss": 0.0089, + "step": 105380 + }, + { + "epoch": 30.663369217340705, + "grad_norm": 0.0879669338464737, + "learning_rate": 1.1995980231137038e-06, + "loss": 0.0116, + "step": 105390 + }, + { + "epoch": 30.666278731451847, + "grad_norm": 0.0995737612247467, + "learning_rate": 1.1979787442791363e-06, + "loss": 0.0121, + "step": 105400 + }, + { + "epoch": 30.669188245562992, + "grad_norm": 0.10677237808704376, + "learning_rate": 1.1963605136169959e-06, + "loss": 0.0109, + "step": 105410 + }, + { + "epoch": 30.672097759674134, + "grad_norm": 0.17473742365837097, + "learning_rate": 1.194743331250176e-06, + "loss": 0.0104, + "step": 105420 + }, + { + "epoch": 30.675007273785276, + "grad_norm": 0.1314888447523117, + "learning_rate": 1.1931271973014952e-06, + "loss": 0.0121, + "step": 105430 + }, + { + "epoch": 30.67791678789642, + "grad_norm": 0.16882877051830292, + "learning_rate": 1.1915121118936866e-06, + "loss": 0.0084, + "step": 105440 + }, + { + "epoch": 30.680826302007564, + "grad_norm": 0.06116114556789398, + "learning_rate": 1.1898980751494043e-06, + "loss": 0.0103, + "step": 105450 + }, + { + "epoch": 30.68373581611871, + "grad_norm": 0.11242896318435669, + "learning_rate": 1.1882850871912231e-06, + "loss": 0.0112, + "step": 105460 + }, + { + "epoch": 30.68664533022985, + "grad_norm": 0.10195382684469223, + "learning_rate": 1.1866731481416438e-06, + "loss": 0.0101, + "step": 105470 + }, + { + "epoch": 30.689554844340996, + "grad_norm": 0.10099457949399948, + "learning_rate": 1.1850622581230747e-06, + "loss": 0.009, + "step": 105480 + }, + { + "epoch": 30.69246435845214, + "grad_norm": 0.12538045644760132, + "learning_rate": 1.1834524172578598e-06, + "loss": 0.0116, + "step": 105490 + }, + { + "epoch": 30.69537387256328, + "grad_norm": 0.1501418501138687, + "learning_rate": 1.1818436256682514e-06, + "loss": 0.01, + "step": 105500 + }, + { + "epoch": 30.698283386674426, + "grad_norm": 0.12243220955133438, + "learning_rate": 1.1802358834764283e-06, + "loss": 0.0117, + "step": 105510 + }, + { + "epoch": 30.701192900785568, + "grad_norm": 0.09787268191576004, + "learning_rate": 1.178629190804486e-06, + "loss": 0.012, + "step": 105520 + }, + { + "epoch": 30.704102414896713, + "grad_norm": 0.16230736672878265, + "learning_rate": 1.1770235477744457e-06, + "loss": 0.0126, + "step": 105530 + }, + { + "epoch": 30.707011929007855, + "grad_norm": 0.11497882753610611, + "learning_rate": 1.1754189545082394e-06, + "loss": 0.0098, + "step": 105540 + }, + { + "epoch": 30.709921443119, + "grad_norm": 0.11968881636857986, + "learning_rate": 1.1738154111277317e-06, + "loss": 0.0086, + "step": 105550 + }, + { + "epoch": 30.712830957230143, + "grad_norm": 0.1798691749572754, + "learning_rate": 1.1722129177546997e-06, + "loss": 0.0099, + "step": 105560 + }, + { + "epoch": 30.715740471341284, + "grad_norm": 0.1273152083158493, + "learning_rate": 1.1706114745108381e-06, + "loss": 0.0113, + "step": 105570 + }, + { + "epoch": 30.71864998545243, + "grad_norm": 0.11380680650472641, + "learning_rate": 1.1690110815177697e-06, + "loss": 0.0105, + "step": 105580 + }, + { + "epoch": 30.721559499563572, + "grad_norm": 0.07267022132873535, + "learning_rate": 1.1674117388970367e-06, + "loss": 0.0103, + "step": 105590 + }, + { + "epoch": 30.724469013674717, + "grad_norm": 0.11756256222724915, + "learning_rate": 1.1658134467700927e-06, + "loss": 0.0111, + "step": 105600 + }, + { + "epoch": 30.72737852778586, + "grad_norm": 0.0957062691450119, + "learning_rate": 1.1642162052583182e-06, + "loss": 0.0097, + "step": 105610 + }, + { + "epoch": 30.730288041897005, + "grad_norm": 0.1110597550868988, + "learning_rate": 1.162620014483015e-06, + "loss": 0.0108, + "step": 105620 + }, + { + "epoch": 30.733197556008147, + "grad_norm": 0.06420084089040756, + "learning_rate": 1.1610248745654061e-06, + "loss": 0.011, + "step": 105630 + }, + { + "epoch": 30.73610707011929, + "grad_norm": 0.08937204629182816, + "learning_rate": 1.1594307856266278e-06, + "loss": 0.0096, + "step": 105640 + }, + { + "epoch": 30.739016584230434, + "grad_norm": 0.08628872781991959, + "learning_rate": 1.1578377477877366e-06, + "loss": 0.0109, + "step": 105650 + }, + { + "epoch": 30.741926098341576, + "grad_norm": 0.1813594251871109, + "learning_rate": 1.1562457611697241e-06, + "loss": 0.0114, + "step": 105660 + }, + { + "epoch": 30.74483561245272, + "grad_norm": 0.1289631873369217, + "learning_rate": 1.1546548258934837e-06, + "loss": 0.0107, + "step": 105670 + }, + { + "epoch": 30.747745126563863, + "grad_norm": 0.12353561818599701, + "learning_rate": 1.1530649420798368e-06, + "loss": 0.0092, + "step": 105680 + }, + { + "epoch": 30.75065464067501, + "grad_norm": 0.10520146787166595, + "learning_rate": 1.1514761098495236e-06, + "loss": 0.0113, + "step": 105690 + }, + { + "epoch": 30.75356415478615, + "grad_norm": 0.0957864448428154, + "learning_rate": 1.1498883293232126e-06, + "loss": 0.0117, + "step": 105700 + }, + { + "epoch": 30.756473668897293, + "grad_norm": 0.10428518056869507, + "learning_rate": 1.1483016006214793e-06, + "loss": 0.009, + "step": 105710 + }, + { + "epoch": 30.75938318300844, + "grad_norm": 0.1630682796239853, + "learning_rate": 1.14671592386482e-06, + "loss": 0.0103, + "step": 105720 + }, + { + "epoch": 30.76229269711958, + "grad_norm": 0.05504153296351433, + "learning_rate": 1.145131299173669e-06, + "loss": 0.0105, + "step": 105730 + }, + { + "epoch": 30.765202211230726, + "grad_norm": 0.15049174427986145, + "learning_rate": 1.1435477266683614e-06, + "loss": 0.0101, + "step": 105740 + }, + { + "epoch": 30.768111725341868, + "grad_norm": 0.1247660219669342, + "learning_rate": 1.1419652064691561e-06, + "loss": 0.01, + "step": 105750 + }, + { + "epoch": 30.771021239453013, + "grad_norm": 0.08731770515441895, + "learning_rate": 1.1403837386962384e-06, + "loss": 0.0093, + "step": 105760 + }, + { + "epoch": 30.773930753564155, + "grad_norm": 0.07357361167669296, + "learning_rate": 1.1388033234697127e-06, + "loss": 0.009, + "step": 105770 + }, + { + "epoch": 30.776840267675297, + "grad_norm": 0.11443887650966644, + "learning_rate": 1.1372239609095992e-06, + "loss": 0.013, + "step": 105780 + }, + { + "epoch": 30.779749781786442, + "grad_norm": 0.09305381029844284, + "learning_rate": 1.135645651135837e-06, + "loss": 0.0096, + "step": 105790 + }, + { + "epoch": 30.782659295897584, + "grad_norm": 0.1310179978609085, + "learning_rate": 1.1340683942682905e-06, + "loss": 0.0104, + "step": 105800 + }, + { + "epoch": 30.78556881000873, + "grad_norm": 0.09404555708169937, + "learning_rate": 1.1324921904267454e-06, + "loss": 0.0097, + "step": 105810 + }, + { + "epoch": 30.788478324119872, + "grad_norm": 0.1702103316783905, + "learning_rate": 1.1309170397308977e-06, + "loss": 0.011, + "step": 105820 + }, + { + "epoch": 30.791387838231014, + "grad_norm": 0.07570072263479233, + "learning_rate": 1.1293429423003732e-06, + "loss": 0.0095, + "step": 105830 + }, + { + "epoch": 30.79429735234216, + "grad_norm": 0.09993129968643188, + "learning_rate": 1.1277698982547162e-06, + "loss": 0.0079, + "step": 105840 + }, + { + "epoch": 30.7972068664533, + "grad_norm": 0.18619219958782196, + "learning_rate": 1.126197907713385e-06, + "loss": 0.0107, + "step": 105850 + }, + { + "epoch": 30.800116380564447, + "grad_norm": 0.0764448493719101, + "learning_rate": 1.1246269707957652e-06, + "loss": 0.0101, + "step": 105860 + }, + { + "epoch": 30.80302589467559, + "grad_norm": 0.1756758689880371, + "learning_rate": 1.1230570876211549e-06, + "loss": 0.0114, + "step": 105870 + }, + { + "epoch": 30.805935408786734, + "grad_norm": 0.1645403951406479, + "learning_rate": 1.1214882583087805e-06, + "loss": 0.011, + "step": 105880 + }, + { + "epoch": 30.808844922897876, + "grad_norm": 0.1125633716583252, + "learning_rate": 1.1199204829777814e-06, + "loss": 0.0108, + "step": 105890 + }, + { + "epoch": 30.811754437009018, + "grad_norm": 0.11223208159208298, + "learning_rate": 1.1183537617472246e-06, + "loss": 0.0106, + "step": 105900 + }, + { + "epoch": 30.814663951120163, + "grad_norm": 0.0834275558590889, + "learning_rate": 1.1167880947360842e-06, + "loss": 0.0107, + "step": 105910 + }, + { + "epoch": 30.817573465231305, + "grad_norm": 0.08721768110990524, + "learning_rate": 1.1152234820632672e-06, + "loss": 0.0116, + "step": 105920 + }, + { + "epoch": 30.82048297934245, + "grad_norm": 0.08290717005729675, + "learning_rate": 1.1136599238475986e-06, + "loss": 0.0105, + "step": 105930 + }, + { + "epoch": 30.823392493453593, + "grad_norm": 0.08422444760799408, + "learning_rate": 1.1120974202078133e-06, + "loss": 0.0098, + "step": 105940 + }, + { + "epoch": 30.82630200756474, + "grad_norm": 0.08187083899974823, + "learning_rate": 1.1105359712625778e-06, + "loss": 0.0098, + "step": 105950 + }, + { + "epoch": 30.82921152167588, + "grad_norm": 0.0955839604139328, + "learning_rate": 1.1089755771304743e-06, + "loss": 0.0098, + "step": 105960 + }, + { + "epoch": 30.832121035787022, + "grad_norm": 0.10452260076999664, + "learning_rate": 1.1074162379300046e-06, + "loss": 0.0106, + "step": 105970 + }, + { + "epoch": 30.835030549898168, + "grad_norm": 0.14107225835323334, + "learning_rate": 1.1058579537795838e-06, + "loss": 0.0112, + "step": 105980 + }, + { + "epoch": 30.83794006400931, + "grad_norm": 0.10060081630945206, + "learning_rate": 1.1043007247975605e-06, + "loss": 0.009, + "step": 105990 + }, + { + "epoch": 30.840849578120455, + "grad_norm": 0.11708849668502808, + "learning_rate": 1.1027445511021939e-06, + "loss": 0.01, + "step": 106000 + }, + { + "epoch": 30.843759092231597, + "grad_norm": 0.14106833934783936, + "learning_rate": 1.1011894328116656e-06, + "loss": 0.0116, + "step": 106010 + }, + { + "epoch": 30.846668606342742, + "grad_norm": 0.1325356811285019, + "learning_rate": 1.0996353700440713e-06, + "loss": 0.0118, + "step": 106020 + }, + { + "epoch": 30.849578120453884, + "grad_norm": 0.142317995429039, + "learning_rate": 1.0980823629174403e-06, + "loss": 0.0111, + "step": 106030 + }, + { + "epoch": 30.852487634565026, + "grad_norm": 0.10578326880931854, + "learning_rate": 1.0965304115497094e-06, + "loss": 0.0111, + "step": 106040 + }, + { + "epoch": 30.855397148676172, + "grad_norm": 0.1172722652554512, + "learning_rate": 1.0949795160587361e-06, + "loss": 0.0113, + "step": 106050 + }, + { + "epoch": 30.858306662787314, + "grad_norm": 0.13601167500019073, + "learning_rate": 1.093429676562303e-06, + "loss": 0.0103, + "step": 106060 + }, + { + "epoch": 30.86121617689846, + "grad_norm": 0.14073127508163452, + "learning_rate": 1.0918808931781144e-06, + "loss": 0.0124, + "step": 106070 + }, + { + "epoch": 30.8641256910096, + "grad_norm": 0.07986371219158173, + "learning_rate": 1.090333166023786e-06, + "loss": 0.011, + "step": 106080 + }, + { + "epoch": 30.867035205120743, + "grad_norm": 0.0914507582783699, + "learning_rate": 1.0887864952168554e-06, + "loss": 0.0133, + "step": 106090 + }, + { + "epoch": 30.86994471923189, + "grad_norm": 0.06820007413625717, + "learning_rate": 1.0872408808747837e-06, + "loss": 0.0102, + "step": 106100 + }, + { + "epoch": 30.87285423334303, + "grad_norm": 0.06841875612735748, + "learning_rate": 1.0856963231149535e-06, + "loss": 0.0101, + "step": 106110 + }, + { + "epoch": 30.875763747454176, + "grad_norm": 0.14162376523017883, + "learning_rate": 1.0841528220546598e-06, + "loss": 0.0124, + "step": 106120 + }, + { + "epoch": 30.878673261565318, + "grad_norm": 0.0794142335653305, + "learning_rate": 1.0826103778111218e-06, + "loss": 0.0104, + "step": 106130 + }, + { + "epoch": 30.881582775676463, + "grad_norm": 0.07862628251314163, + "learning_rate": 1.0810689905014808e-06, + "loss": 0.0102, + "step": 106140 + }, + { + "epoch": 30.884492289787605, + "grad_norm": 0.11442065984010696, + "learning_rate": 1.0795286602427933e-06, + "loss": 0.0108, + "step": 106150 + }, + { + "epoch": 30.887401803898747, + "grad_norm": 0.10009769350290298, + "learning_rate": 1.0779893871520342e-06, + "loss": 0.0111, + "step": 106160 + }, + { + "epoch": 30.890311318009893, + "grad_norm": 0.1168542355298996, + "learning_rate": 1.0764511713461044e-06, + "loss": 0.0098, + "step": 106170 + }, + { + "epoch": 30.893220832121035, + "grad_norm": 0.18982963263988495, + "learning_rate": 1.0749140129418228e-06, + "loss": 0.0118, + "step": 106180 + }, + { + "epoch": 30.89613034623218, + "grad_norm": 0.09686245024204254, + "learning_rate": 1.0733779120559223e-06, + "loss": 0.009, + "step": 106190 + }, + { + "epoch": 30.899039860343322, + "grad_norm": 0.17040592432022095, + "learning_rate": 1.0718428688050647e-06, + "loss": 0.0118, + "step": 106200 + }, + { + "epoch": 30.901949374454468, + "grad_norm": 0.08529935777187347, + "learning_rate": 1.07030888330582e-06, + "loss": 0.01, + "step": 106210 + }, + { + "epoch": 30.90485888856561, + "grad_norm": 0.08268829435110092, + "learning_rate": 1.0687759556746884e-06, + "loss": 0.0089, + "step": 106220 + }, + { + "epoch": 30.90776840267675, + "grad_norm": 0.08449345082044601, + "learning_rate": 1.067244086028087e-06, + "loss": 0.0098, + "step": 106230 + }, + { + "epoch": 30.910677916787897, + "grad_norm": 0.09079550206661224, + "learning_rate": 1.0657132744823456e-06, + "loss": 0.0083, + "step": 106240 + }, + { + "epoch": 30.91358743089904, + "grad_norm": 0.129109188914299, + "learning_rate": 1.0641835211537265e-06, + "loss": 0.0081, + "step": 106250 + }, + { + "epoch": 30.916496945010184, + "grad_norm": 0.10426995903253555, + "learning_rate": 1.0626548261583986e-06, + "loss": 0.0102, + "step": 106260 + }, + { + "epoch": 30.919406459121326, + "grad_norm": 0.08247467875480652, + "learning_rate": 1.0611271896124585e-06, + "loss": 0.009, + "step": 106270 + }, + { + "epoch": 30.92231597323247, + "grad_norm": 0.08955986052751541, + "learning_rate": 1.059600611631919e-06, + "loss": 0.0102, + "step": 106280 + }, + { + "epoch": 30.925225487343614, + "grad_norm": 0.20500734448432922, + "learning_rate": 1.0580750923327136e-06, + "loss": 0.0123, + "step": 106290 + }, + { + "epoch": 30.928135001454756, + "grad_norm": 0.10324487835168839, + "learning_rate": 1.0565506318307e-06, + "loss": 0.0104, + "step": 106300 + }, + { + "epoch": 30.9310445155659, + "grad_norm": 0.20457687973976135, + "learning_rate": 1.0550272302416453e-06, + "loss": 0.0091, + "step": 106310 + }, + { + "epoch": 30.933954029677043, + "grad_norm": 0.15484948456287384, + "learning_rate": 1.0535048876812402e-06, + "loss": 0.0096, + "step": 106320 + }, + { + "epoch": 30.93686354378819, + "grad_norm": 0.0992451161146164, + "learning_rate": 1.0519836042651042e-06, + "loss": 0.0117, + "step": 106330 + }, + { + "epoch": 30.93977305789933, + "grad_norm": 0.11159677803516388, + "learning_rate": 1.0504633801087648e-06, + "loss": 0.0099, + "step": 106340 + }, + { + "epoch": 30.942682572010476, + "grad_norm": 0.09494244307279587, + "learning_rate": 1.0489442153276696e-06, + "loss": 0.0088, + "step": 106350 + }, + { + "epoch": 30.945592086121618, + "grad_norm": 0.11522167921066284, + "learning_rate": 1.047426110037193e-06, + "loss": 0.0098, + "step": 106360 + }, + { + "epoch": 30.94850160023276, + "grad_norm": 0.1437743902206421, + "learning_rate": 1.0459090643526275e-06, + "loss": 0.0104, + "step": 106370 + }, + { + "epoch": 30.951411114343905, + "grad_norm": 0.11529731750488281, + "learning_rate": 1.0443930783891782e-06, + "loss": 0.0142, + "step": 106380 + }, + { + "epoch": 30.954320628455047, + "grad_norm": 0.07113231718540192, + "learning_rate": 1.0428781522619741e-06, + "loss": 0.0093, + "step": 106390 + }, + { + "epoch": 30.957230142566193, + "grad_norm": 0.1055554449558258, + "learning_rate": 1.041364286086065e-06, + "loss": 0.0088, + "step": 106400 + }, + { + "epoch": 30.960139656677335, + "grad_norm": 0.08658762276172638, + "learning_rate": 1.0398514799764226e-06, + "loss": 0.0094, + "step": 106410 + }, + { + "epoch": 30.96304917078848, + "grad_norm": 0.11672456562519073, + "learning_rate": 1.0383397340479279e-06, + "loss": 0.0109, + "step": 106420 + }, + { + "epoch": 30.965958684899622, + "grad_norm": 0.1820676177740097, + "learning_rate": 1.0368290484153908e-06, + "loss": 0.0101, + "step": 106430 + }, + { + "epoch": 30.968868199010764, + "grad_norm": 0.12512712180614471, + "learning_rate": 1.0353194231935431e-06, + "loss": 0.0122, + "step": 106440 + }, + { + "epoch": 30.97177771312191, + "grad_norm": 0.18369877338409424, + "learning_rate": 1.0338108584970247e-06, + "loss": 0.0099, + "step": 106450 + }, + { + "epoch": 30.97468722723305, + "grad_norm": 0.09766913205385208, + "learning_rate": 1.0323033544404003e-06, + "loss": 0.0116, + "step": 106460 + }, + { + "epoch": 30.977596741344197, + "grad_norm": 0.0816168487071991, + "learning_rate": 1.0307969111381588e-06, + "loss": 0.0112, + "step": 106470 + }, + { + "epoch": 30.98050625545534, + "grad_norm": 0.09640775620937347, + "learning_rate": 1.0292915287047033e-06, + "loss": 0.01, + "step": 106480 + }, + { + "epoch": 30.98341576956648, + "grad_norm": 0.16401654481887817, + "learning_rate": 1.0277872072543576e-06, + "loss": 0.0129, + "step": 106490 + }, + { + "epoch": 30.986325283677626, + "grad_norm": 0.0927242636680603, + "learning_rate": 1.0262839469013653e-06, + "loss": 0.0089, + "step": 106500 + }, + { + "epoch": 30.989234797788768, + "grad_norm": 0.11998310685157776, + "learning_rate": 1.0247817477598882e-06, + "loss": 0.0112, + "step": 106510 + }, + { + "epoch": 30.992144311899914, + "grad_norm": 0.09637358039617538, + "learning_rate": 1.02328060994401e-06, + "loss": 0.0116, + "step": 106520 + }, + { + "epoch": 30.995053826011056, + "grad_norm": 0.12663063406944275, + "learning_rate": 1.0217805335677282e-06, + "loss": 0.0116, + "step": 106530 + }, + { + "epoch": 30.9979633401222, + "grad_norm": 0.10349169373512268, + "learning_rate": 1.0202815187449683e-06, + "loss": 0.0104, + "step": 106540 + }, + { + "epoch": 31.000872854233343, + "grad_norm": 0.08579937368631363, + "learning_rate": 1.0187835655895706e-06, + "loss": 0.0105, + "step": 106550 + }, + { + "epoch": 31.003782368344485, + "grad_norm": 0.11744187027215958, + "learning_rate": 1.017286674215291e-06, + "loss": 0.0096, + "step": 106560 + }, + { + "epoch": 31.00669188245563, + "grad_norm": 0.06384608149528503, + "learning_rate": 1.0157908447358117e-06, + "loss": 0.0093, + "step": 106570 + }, + { + "epoch": 31.009601396566772, + "grad_norm": 0.11971959471702576, + "learning_rate": 1.0142960772647285e-06, + "loss": 0.0084, + "step": 106580 + }, + { + "epoch": 31.012510910677918, + "grad_norm": 0.17587754130363464, + "learning_rate": 1.012802371915561e-06, + "loss": 0.0086, + "step": 106590 + }, + { + "epoch": 31.01542042478906, + "grad_norm": 0.07149980217218399, + "learning_rate": 1.0113097288017481e-06, + "loss": 0.0088, + "step": 106600 + }, + { + "epoch": 31.018329938900205, + "grad_norm": 0.07799102365970612, + "learning_rate": 1.0098181480366442e-06, + "loss": 0.0096, + "step": 106610 + }, + { + "epoch": 31.021239453011347, + "grad_norm": 0.10590581595897675, + "learning_rate": 1.008327629733522e-06, + "loss": 0.0094, + "step": 106620 + }, + { + "epoch": 31.02414896712249, + "grad_norm": 0.13652969896793365, + "learning_rate": 1.0068381740055792e-06, + "loss": 0.0104, + "step": 106630 + }, + { + "epoch": 31.027058481233635, + "grad_norm": 0.065923310816288, + "learning_rate": 1.005349780965934e-06, + "loss": 0.0114, + "step": 106640 + }, + { + "epoch": 31.029967995344776, + "grad_norm": 0.14813199639320374, + "learning_rate": 1.0038624507276123e-06, + "loss": 0.0124, + "step": 106650 + }, + { + "epoch": 31.032877509455922, + "grad_norm": 0.10884464532136917, + "learning_rate": 1.0023761834035706e-06, + "loss": 0.0107, + "step": 106660 + }, + { + "epoch": 31.035787023567064, + "grad_norm": 0.11700153350830078, + "learning_rate": 1.000890979106685e-06, + "loss": 0.0118, + "step": 106670 + }, + { + "epoch": 31.03869653767821, + "grad_norm": 0.12013102322816849, + "learning_rate": 9.99406837949744e-07, + "loss": 0.0104, + "step": 106680 + }, + { + "epoch": 31.04160605178935, + "grad_norm": 0.11013998836278915, + "learning_rate": 9.97923760045454e-07, + "loss": 0.0114, + "step": 106690 + }, + { + "epoch": 31.044515565900493, + "grad_norm": 0.10817331075668335, + "learning_rate": 9.964417455064495e-07, + "loss": 0.0116, + "step": 106700 + }, + { + "epoch": 31.04742508001164, + "grad_norm": 0.08787331730127335, + "learning_rate": 9.94960794445281e-07, + "loss": 0.0118, + "step": 106710 + }, + { + "epoch": 31.05033459412278, + "grad_norm": 0.0921345129609108, + "learning_rate": 9.93480906974415e-07, + "loss": 0.0113, + "step": 106720 + }, + { + "epoch": 31.053244108233926, + "grad_norm": 0.12044978141784668, + "learning_rate": 9.920020832062348e-07, + "loss": 0.0115, + "step": 106730 + }, + { + "epoch": 31.056153622345068, + "grad_norm": 0.09098994731903076, + "learning_rate": 9.905243232530559e-07, + "loss": 0.0109, + "step": 106740 + }, + { + "epoch": 31.059063136456214, + "grad_norm": 0.11655297875404358, + "learning_rate": 9.890476272270998e-07, + "loss": 0.0106, + "step": 106750 + }, + { + "epoch": 31.061972650567355, + "grad_norm": 0.09872806072235107, + "learning_rate": 9.875719952405104e-07, + "loss": 0.0101, + "step": 106760 + }, + { + "epoch": 31.064882164678497, + "grad_norm": 0.13221772015094757, + "learning_rate": 9.860974274053547e-07, + "loss": 0.0113, + "step": 106770 + }, + { + "epoch": 31.067791678789643, + "grad_norm": 0.08690030127763748, + "learning_rate": 9.846239238336163e-07, + "loss": 0.0111, + "step": 106780 + }, + { + "epoch": 31.070701192900785, + "grad_norm": 0.1728973686695099, + "learning_rate": 9.831514846371974e-07, + "loss": 0.0095, + "step": 106790 + }, + { + "epoch": 31.07361070701193, + "grad_norm": 0.0826890766620636, + "learning_rate": 9.816801099279204e-07, + "loss": 0.01, + "step": 106800 + }, + { + "epoch": 31.076520221123072, + "grad_norm": 0.08644115179777145, + "learning_rate": 9.80209799817524e-07, + "loss": 0.0099, + "step": 106810 + }, + { + "epoch": 31.079429735234214, + "grad_norm": 0.056658245623111725, + "learning_rate": 9.787405544176743e-07, + "loss": 0.0098, + "step": 106820 + }, + { + "epoch": 31.08233924934536, + "grad_norm": 0.08252207189798355, + "learning_rate": 9.772723738399448e-07, + "loss": 0.008, + "step": 106830 + }, + { + "epoch": 31.0852487634565, + "grad_norm": 0.16919498145580292, + "learning_rate": 9.75805258195835e-07, + "loss": 0.0104, + "step": 106840 + }, + { + "epoch": 31.088158277567647, + "grad_norm": 0.08135602623224258, + "learning_rate": 9.743392075967688e-07, + "loss": 0.012, + "step": 106850 + }, + { + "epoch": 31.09106779167879, + "grad_norm": 0.08779370039701462, + "learning_rate": 9.728742221540744e-07, + "loss": 0.0101, + "step": 106860 + }, + { + "epoch": 31.093977305789934, + "grad_norm": 0.13295549154281616, + "learning_rate": 9.714103019790137e-07, + "loss": 0.0105, + "step": 106870 + }, + { + "epoch": 31.096886819901076, + "grad_norm": 0.09053894877433777, + "learning_rate": 9.699474471827585e-07, + "loss": 0.0111, + "step": 106880 + }, + { + "epoch": 31.09979633401222, + "grad_norm": 0.058107681572437286, + "learning_rate": 9.68485657876408e-07, + "loss": 0.0108, + "step": 106890 + }, + { + "epoch": 31.102705848123364, + "grad_norm": 0.08478999882936478, + "learning_rate": 9.670249341709681e-07, + "loss": 0.0124, + "step": 106900 + }, + { + "epoch": 31.105615362234506, + "grad_norm": 0.10562200099229813, + "learning_rate": 9.65565276177377e-07, + "loss": 0.0119, + "step": 106910 + }, + { + "epoch": 31.10852487634565, + "grad_norm": 0.09630890190601349, + "learning_rate": 9.641066840064827e-07, + "loss": 0.0105, + "step": 106920 + }, + { + "epoch": 31.111434390456793, + "grad_norm": 0.146344855427742, + "learning_rate": 9.62649157769055e-07, + "loss": 0.0115, + "step": 106930 + }, + { + "epoch": 31.11434390456794, + "grad_norm": 0.1401563435792923, + "learning_rate": 9.6119269757579e-07, + "loss": 0.0093, + "step": 106940 + }, + { + "epoch": 31.11725341867908, + "grad_norm": 0.11713916808366776, + "learning_rate": 9.59737303537288e-07, + "loss": 0.0112, + "step": 106950 + }, + { + "epoch": 31.120162932790222, + "grad_norm": 0.10461767017841339, + "learning_rate": 9.582829757640804e-07, + "loss": 0.0091, + "step": 106960 + }, + { + "epoch": 31.123072446901368, + "grad_norm": 0.11929894983768463, + "learning_rate": 9.568297143666154e-07, + "loss": 0.0102, + "step": 106970 + }, + { + "epoch": 31.12598196101251, + "grad_norm": 0.16574062407016754, + "learning_rate": 9.553775194552566e-07, + "loss": 0.0108, + "step": 106980 + }, + { + "epoch": 31.128891475123655, + "grad_norm": 0.0890357717871666, + "learning_rate": 9.539263911402873e-07, + "loss": 0.0112, + "step": 106990 + }, + { + "epoch": 31.131800989234797, + "grad_norm": 0.08795329928398132, + "learning_rate": 9.524763295319127e-07, + "loss": 0.0094, + "step": 107000 + }, + { + "epoch": 31.134710503345943, + "grad_norm": 0.09672216325998306, + "learning_rate": 9.510273347402565e-07, + "loss": 0.0132, + "step": 107010 + }, + { + "epoch": 31.137620017457085, + "grad_norm": 0.12338461726903915, + "learning_rate": 9.495794068753588e-07, + "loss": 0.0108, + "step": 107020 + }, + { + "epoch": 31.140529531568227, + "grad_norm": 0.07507596164941788, + "learning_rate": 9.481325460471769e-07, + "loss": 0.0094, + "step": 107030 + }, + { + "epoch": 31.143439045679372, + "grad_norm": 0.06890665739774704, + "learning_rate": 9.466867523655992e-07, + "loss": 0.0124, + "step": 107040 + }, + { + "epoch": 31.146348559790514, + "grad_norm": 0.1802268773317337, + "learning_rate": 9.452420259404166e-07, + "loss": 0.0107, + "step": 107050 + }, + { + "epoch": 31.14925807390166, + "grad_norm": 0.1503475159406662, + "learning_rate": 9.437983668813477e-07, + "loss": 0.0097, + "step": 107060 + }, + { + "epoch": 31.1521675880128, + "grad_norm": 0.07442568242549896, + "learning_rate": 9.4235577529803e-07, + "loss": 0.0104, + "step": 107070 + }, + { + "epoch": 31.155077102123947, + "grad_norm": 0.2562362253665924, + "learning_rate": 9.409142513000224e-07, + "loss": 0.0097, + "step": 107080 + }, + { + "epoch": 31.15798661623509, + "grad_norm": 0.11122303456068039, + "learning_rate": 9.394737949967941e-07, + "loss": 0.0122, + "step": 107090 + }, + { + "epoch": 31.16089613034623, + "grad_norm": 0.11476790904998779, + "learning_rate": 9.380344064977376e-07, + "loss": 0.0111, + "step": 107100 + }, + { + "epoch": 31.163805644457376, + "grad_norm": 0.08415137976408005, + "learning_rate": 9.36596085912167e-07, + "loss": 0.0085, + "step": 107110 + }, + { + "epoch": 31.16671515856852, + "grad_norm": 0.11105503886938095, + "learning_rate": 9.351588333493149e-07, + "loss": 0.0088, + "step": 107120 + }, + { + "epoch": 31.169624672679664, + "grad_norm": 0.08115503937005997, + "learning_rate": 9.337226489183275e-07, + "loss": 0.0089, + "step": 107130 + }, + { + "epoch": 31.172534186790806, + "grad_norm": 0.0769832581281662, + "learning_rate": 9.322875327282759e-07, + "loss": 0.0115, + "step": 107140 + }, + { + "epoch": 31.17544370090195, + "grad_norm": 0.207894429564476, + "learning_rate": 9.308534848881495e-07, + "loss": 0.0111, + "step": 107150 + }, + { + "epoch": 31.178353215013093, + "grad_norm": 0.08494050800800323, + "learning_rate": 9.294205055068511e-07, + "loss": 0.0103, + "step": 107160 + }, + { + "epoch": 31.181262729124235, + "grad_norm": 0.0958627387881279, + "learning_rate": 9.279885946932088e-07, + "loss": 0.0098, + "step": 107170 + }, + { + "epoch": 31.18417224323538, + "grad_norm": 0.07292490452528, + "learning_rate": 9.265577525559637e-07, + "loss": 0.0103, + "step": 107180 + }, + { + "epoch": 31.187081757346522, + "grad_norm": 0.10469215363264084, + "learning_rate": 9.251279792037843e-07, + "loss": 0.0091, + "step": 107190 + }, + { + "epoch": 31.189991271457668, + "grad_norm": 0.1063617616891861, + "learning_rate": 9.23699274745245e-07, + "loss": 0.0113, + "step": 107200 + }, + { + "epoch": 31.19290078556881, + "grad_norm": 0.10038509219884872, + "learning_rate": 9.222716392888525e-07, + "loss": 0.0096, + "step": 107210 + }, + { + "epoch": 31.195810299679952, + "grad_norm": 0.1481461077928543, + "learning_rate": 9.208450729430217e-07, + "loss": 0.0134, + "step": 107220 + }, + { + "epoch": 31.198719813791097, + "grad_norm": 0.08617675304412842, + "learning_rate": 9.194195758160944e-07, + "loss": 0.0109, + "step": 107230 + }, + { + "epoch": 31.20162932790224, + "grad_norm": 0.07041136920452118, + "learning_rate": 9.179951480163273e-07, + "loss": 0.0094, + "step": 107240 + }, + { + "epoch": 31.204538842013385, + "grad_norm": 0.0943809300661087, + "learning_rate": 9.165717896518938e-07, + "loss": 0.0091, + "step": 107250 + }, + { + "epoch": 31.207448356124527, + "grad_norm": 0.1798590123653412, + "learning_rate": 9.151495008308941e-07, + "loss": 0.011, + "step": 107260 + }, + { + "epoch": 31.210357870235672, + "grad_norm": 0.12950661778450012, + "learning_rate": 9.137282816613335e-07, + "loss": 0.0117, + "step": 107270 + }, + { + "epoch": 31.213267384346814, + "grad_norm": 0.10606164485216141, + "learning_rate": 9.123081322511523e-07, + "loss": 0.0093, + "step": 107280 + }, + { + "epoch": 31.216176898457956, + "grad_norm": 0.0800674632191658, + "learning_rate": 9.108890527081942e-07, + "loss": 0.0097, + "step": 107290 + }, + { + "epoch": 31.2190864125691, + "grad_norm": 0.0868753120303154, + "learning_rate": 9.094710431402331e-07, + "loss": 0.0094, + "step": 107300 + }, + { + "epoch": 31.221995926680243, + "grad_norm": 0.1068207398056984, + "learning_rate": 9.080541036549594e-07, + "loss": 0.0118, + "step": 107310 + }, + { + "epoch": 31.22490544079139, + "grad_norm": 0.10670715570449829, + "learning_rate": 9.066382343599789e-07, + "loss": 0.011, + "step": 107320 + }, + { + "epoch": 31.22781495490253, + "grad_norm": 0.12495718896389008, + "learning_rate": 9.052234353628103e-07, + "loss": 0.0122, + "step": 107330 + }, + { + "epoch": 31.230724469013676, + "grad_norm": 0.15663863718509674, + "learning_rate": 9.038097067709094e-07, + "loss": 0.0092, + "step": 107340 + }, + { + "epoch": 31.233633983124818, + "grad_norm": 0.08742251992225647, + "learning_rate": 9.023970486916339e-07, + "loss": 0.0115, + "step": 107350 + }, + { + "epoch": 31.23654349723596, + "grad_norm": 0.11078950762748718, + "learning_rate": 9.009854612322643e-07, + "loss": 0.0111, + "step": 107360 + }, + { + "epoch": 31.239453011347106, + "grad_norm": 0.12824976444244385, + "learning_rate": 8.995749445000051e-07, + "loss": 0.011, + "step": 107370 + }, + { + "epoch": 31.242362525458248, + "grad_norm": 0.10257257521152496, + "learning_rate": 8.981654986019739e-07, + "loss": 0.0092, + "step": 107380 + }, + { + "epoch": 31.245272039569393, + "grad_norm": 0.12422682344913483, + "learning_rate": 8.967571236452116e-07, + "loss": 0.0098, + "step": 107390 + }, + { + "epoch": 31.248181553680535, + "grad_norm": 0.13038231432437897, + "learning_rate": 8.953498197366677e-07, + "loss": 0.0093, + "step": 107400 + }, + { + "epoch": 31.25109106779168, + "grad_norm": 0.11805083602666855, + "learning_rate": 8.939435869832235e-07, + "loss": 0.0087, + "step": 107410 + }, + { + "epoch": 31.254000581902822, + "grad_norm": 0.12590573728084564, + "learning_rate": 8.925384254916752e-07, + "loss": 0.0135, + "step": 107420 + }, + { + "epoch": 31.256910096013964, + "grad_norm": 0.09681171178817749, + "learning_rate": 8.91134335368729e-07, + "loss": 0.0095, + "step": 107430 + }, + { + "epoch": 31.25981961012511, + "grad_norm": 0.08039841800928116, + "learning_rate": 8.897313167210197e-07, + "loss": 0.0097, + "step": 107440 + }, + { + "epoch": 31.26272912423625, + "grad_norm": 0.08205942809581757, + "learning_rate": 8.883293696551003e-07, + "loss": 0.0084, + "step": 107450 + }, + { + "epoch": 31.265638638347397, + "grad_norm": 0.13742157816886902, + "learning_rate": 8.869284942774359e-07, + "loss": 0.0111, + "step": 107460 + }, + { + "epoch": 31.26854815245854, + "grad_norm": 0.07365301996469498, + "learning_rate": 8.85528690694411e-07, + "loss": 0.0107, + "step": 107470 + }, + { + "epoch": 31.27145766656968, + "grad_norm": 0.11580245941877365, + "learning_rate": 8.841299590123358e-07, + "loss": 0.0122, + "step": 107480 + }, + { + "epoch": 31.274367180680827, + "grad_norm": 0.1585770696401596, + "learning_rate": 8.827322993374354e-07, + "loss": 0.0095, + "step": 107490 + }, + { + "epoch": 31.27727669479197, + "grad_norm": 0.1696770042181015, + "learning_rate": 8.813357117758497e-07, + "loss": 0.0115, + "step": 107500 + }, + { + "epoch": 31.280186208903114, + "grad_norm": 0.13954579830169678, + "learning_rate": 8.799401964336423e-07, + "loss": 0.009, + "step": 107510 + }, + { + "epoch": 31.283095723014256, + "grad_norm": 0.13162095844745636, + "learning_rate": 8.785457534167918e-07, + "loss": 0.01, + "step": 107520 + }, + { + "epoch": 31.2860052371254, + "grad_norm": 0.13236530125141144, + "learning_rate": 8.771523828311967e-07, + "loss": 0.009, + "step": 107530 + }, + { + "epoch": 31.288914751236543, + "grad_norm": 0.18253971636295319, + "learning_rate": 8.757600847826775e-07, + "loss": 0.0102, + "step": 107540 + }, + { + "epoch": 31.291824265347685, + "grad_norm": 0.09206025302410126, + "learning_rate": 8.743688593769661e-07, + "loss": 0.0099, + "step": 107550 + }, + { + "epoch": 31.29473377945883, + "grad_norm": 0.10042881220579147, + "learning_rate": 8.729787067197215e-07, + "loss": 0.0107, + "step": 107560 + }, + { + "epoch": 31.297643293569973, + "grad_norm": 0.10960332304239273, + "learning_rate": 8.715896269165124e-07, + "loss": 0.0094, + "step": 107570 + }, + { + "epoch": 31.300552807681118, + "grad_norm": 0.14538423717021942, + "learning_rate": 8.70201620072833e-07, + "loss": 0.0121, + "step": 107580 + }, + { + "epoch": 31.30346232179226, + "grad_norm": 0.0609954297542572, + "learning_rate": 8.688146862940887e-07, + "loss": 0.0097, + "step": 107590 + }, + { + "epoch": 31.306371835903406, + "grad_norm": 0.08499401062726974, + "learning_rate": 8.674288256856139e-07, + "loss": 0.0116, + "step": 107600 + }, + { + "epoch": 31.309281350014547, + "grad_norm": 0.07817826420068741, + "learning_rate": 8.660440383526525e-07, + "loss": 0.011, + "step": 107610 + }, + { + "epoch": 31.31219086412569, + "grad_norm": 0.05835286155343056, + "learning_rate": 8.646603244003721e-07, + "loss": 0.0114, + "step": 107620 + }, + { + "epoch": 31.315100378236835, + "grad_norm": 0.12541556358337402, + "learning_rate": 8.632776839338502e-07, + "loss": 0.0101, + "step": 107630 + }, + { + "epoch": 31.318009892347977, + "grad_norm": 0.11655835807323456, + "learning_rate": 8.61896117058098e-07, + "loss": 0.0109, + "step": 107640 + }, + { + "epoch": 31.320919406459122, + "grad_norm": 0.09512688219547272, + "learning_rate": 8.605156238780315e-07, + "loss": 0.0113, + "step": 107650 + }, + { + "epoch": 31.323828920570264, + "grad_norm": 0.08537497371435165, + "learning_rate": 8.591362044984902e-07, + "loss": 0.0105, + "step": 107660 + }, + { + "epoch": 31.32673843468141, + "grad_norm": 0.24676743149757385, + "learning_rate": 8.577578590242319e-07, + "loss": 0.0098, + "step": 107670 + }, + { + "epoch": 31.32964794879255, + "grad_norm": 0.08566449582576752, + "learning_rate": 8.563805875599362e-07, + "loss": 0.0092, + "step": 107680 + }, + { + "epoch": 31.332557462903694, + "grad_norm": 0.11563721299171448, + "learning_rate": 8.550043902101945e-07, + "loss": 0.0098, + "step": 107690 + }, + { + "epoch": 31.33546697701484, + "grad_norm": 0.10295818001031876, + "learning_rate": 8.536292670795181e-07, + "loss": 0.008, + "step": 107700 + }, + { + "epoch": 31.33837649112598, + "grad_norm": 0.09013421088457108, + "learning_rate": 8.52255218272342e-07, + "loss": 0.0129, + "step": 107710 + }, + { + "epoch": 31.341286005237126, + "grad_norm": 0.11347890645265579, + "learning_rate": 8.508822438930175e-07, + "loss": 0.0132, + "step": 107720 + }, + { + "epoch": 31.34419551934827, + "grad_norm": 0.1449856460094452, + "learning_rate": 8.495103440458097e-07, + "loss": 0.0104, + "step": 107730 + }, + { + "epoch": 31.347105033459414, + "grad_norm": 0.14865034818649292, + "learning_rate": 8.481395188349017e-07, + "loss": 0.0093, + "step": 107740 + }, + { + "epoch": 31.350014547570556, + "grad_norm": 0.08852461725473404, + "learning_rate": 8.467697683644071e-07, + "loss": 0.0117, + "step": 107750 + }, + { + "epoch": 31.352924061681698, + "grad_norm": 0.08102378994226456, + "learning_rate": 8.454010927383459e-07, + "loss": 0.0096, + "step": 107760 + }, + { + "epoch": 31.355833575792843, + "grad_norm": 0.13631099462509155, + "learning_rate": 8.440334920606568e-07, + "loss": 0.009, + "step": 107770 + }, + { + "epoch": 31.358743089903985, + "grad_norm": 0.08329478651285172, + "learning_rate": 8.426669664352032e-07, + "loss": 0.0097, + "step": 107780 + }, + { + "epoch": 31.36165260401513, + "grad_norm": 0.10765507072210312, + "learning_rate": 8.413015159657656e-07, + "loss": 0.0113, + "step": 107790 + }, + { + "epoch": 31.364562118126273, + "grad_norm": 0.10821696370840073, + "learning_rate": 8.399371407560358e-07, + "loss": 0.0102, + "step": 107800 + }, + { + "epoch": 31.367471632237418, + "grad_norm": 0.12507161498069763, + "learning_rate": 8.385738409096344e-07, + "loss": 0.0089, + "step": 107810 + }, + { + "epoch": 31.37038114634856, + "grad_norm": 0.16274262964725494, + "learning_rate": 8.37211616530092e-07, + "loss": 0.0091, + "step": 107820 + }, + { + "epoch": 31.373290660459702, + "grad_norm": 0.12736724317073822, + "learning_rate": 8.358504677208606e-07, + "loss": 0.0112, + "step": 107830 + }, + { + "epoch": 31.376200174570847, + "grad_norm": 0.09073193371295929, + "learning_rate": 8.344903945853094e-07, + "loss": 0.0104, + "step": 107840 + }, + { + "epoch": 31.37910968868199, + "grad_norm": 0.1007775366306305, + "learning_rate": 8.33131397226729e-07, + "loss": 0.0102, + "step": 107850 + }, + { + "epoch": 31.382019202793135, + "grad_norm": 0.11953368037939072, + "learning_rate": 8.317734757483287e-07, + "loss": 0.012, + "step": 107860 + }, + { + "epoch": 31.384928716904277, + "grad_norm": 0.12205745279788971, + "learning_rate": 8.304166302532274e-07, + "loss": 0.0112, + "step": 107870 + }, + { + "epoch": 31.38783823101542, + "grad_norm": 0.08809947222471237, + "learning_rate": 8.290608608444744e-07, + "loss": 0.0107, + "step": 107880 + }, + { + "epoch": 31.390747745126564, + "grad_norm": 0.09555160254240036, + "learning_rate": 8.277061676250275e-07, + "loss": 0.0106, + "step": 107890 + }, + { + "epoch": 31.393657259237706, + "grad_norm": 0.11543082445859909, + "learning_rate": 8.263525506977659e-07, + "loss": 0.0106, + "step": 107900 + }, + { + "epoch": 31.39656677334885, + "grad_norm": 0.11570753157138824, + "learning_rate": 8.250000101654942e-07, + "loss": 0.01, + "step": 107910 + }, + { + "epoch": 31.399476287459994, + "grad_norm": 0.16525398194789886, + "learning_rate": 8.236485461309251e-07, + "loss": 0.0117, + "step": 107920 + }, + { + "epoch": 31.40238580157114, + "grad_norm": 0.19606074690818787, + "learning_rate": 8.222981586966899e-07, + "loss": 0.0104, + "step": 107930 + }, + { + "epoch": 31.40529531568228, + "grad_norm": 0.11574406176805496, + "learning_rate": 8.209488479653465e-07, + "loss": 0.0096, + "step": 107940 + }, + { + "epoch": 31.408204829793423, + "grad_norm": 0.08458445966243744, + "learning_rate": 8.196006140393648e-07, + "loss": 0.0111, + "step": 107950 + }, + { + "epoch": 31.41111434390457, + "grad_norm": 0.07316742837429047, + "learning_rate": 8.182534570211325e-07, + "loss": 0.0091, + "step": 107960 + }, + { + "epoch": 31.41402385801571, + "grad_norm": 0.11184930801391602, + "learning_rate": 8.169073770129582e-07, + "loss": 0.0098, + "step": 107970 + }, + { + "epoch": 31.416933372126856, + "grad_norm": 0.09510987997055054, + "learning_rate": 8.155623741170715e-07, + "loss": 0.012, + "step": 107980 + }, + { + "epoch": 31.419842886237998, + "grad_norm": 0.10109162330627441, + "learning_rate": 8.142184484356141e-07, + "loss": 0.0121, + "step": 107990 + }, + { + "epoch": 31.422752400349143, + "grad_norm": 0.12058918178081512, + "learning_rate": 8.128756000706427e-07, + "loss": 0.0103, + "step": 108000 + }, + { + "epoch": 31.425661914460285, + "grad_norm": 0.07423493266105652, + "learning_rate": 8.115338291241442e-07, + "loss": 0.0111, + "step": 108010 + }, + { + "epoch": 31.428571428571427, + "grad_norm": 0.2692904770374298, + "learning_rate": 8.101931356980185e-07, + "loss": 0.0104, + "step": 108020 + }, + { + "epoch": 31.431480942682573, + "grad_norm": 0.10104429721832275, + "learning_rate": 8.088535198940794e-07, + "loss": 0.0097, + "step": 108030 + }, + { + "epoch": 31.434390456793714, + "grad_norm": 0.1152147427201271, + "learning_rate": 8.07514981814057e-07, + "loss": 0.0095, + "step": 108040 + }, + { + "epoch": 31.43729997090486, + "grad_norm": 0.09975415468215942, + "learning_rate": 8.061775215596135e-07, + "loss": 0.0132, + "step": 108050 + }, + { + "epoch": 31.440209485016002, + "grad_norm": 0.12597370147705078, + "learning_rate": 8.048411392323157e-07, + "loss": 0.0107, + "step": 108060 + }, + { + "epoch": 31.443118999127147, + "grad_norm": 0.07578849792480469, + "learning_rate": 8.035058349336527e-07, + "loss": 0.0112, + "step": 108070 + }, + { + "epoch": 31.44602851323829, + "grad_norm": 0.07084766030311584, + "learning_rate": 8.021716087650332e-07, + "loss": 0.0101, + "step": 108080 + }, + { + "epoch": 31.44893802734943, + "grad_norm": 0.14158760011196136, + "learning_rate": 8.008384608277831e-07, + "loss": 0.0101, + "step": 108090 + }, + { + "epoch": 31.451847541460577, + "grad_norm": 0.07354629784822464, + "learning_rate": 7.99506391223146e-07, + "loss": 0.0103, + "step": 108100 + }, + { + "epoch": 31.45475705557172, + "grad_norm": 0.059892114251852036, + "learning_rate": 7.981754000522812e-07, + "loss": 0.009, + "step": 108110 + }, + { + "epoch": 31.457666569682864, + "grad_norm": 0.1404399573802948, + "learning_rate": 7.968454874162711e-07, + "loss": 0.0104, + "step": 108120 + }, + { + "epoch": 31.460576083794006, + "grad_norm": 0.10901843756437302, + "learning_rate": 7.955166534161168e-07, + "loss": 0.0086, + "step": 108130 + }, + { + "epoch": 31.46348559790515, + "grad_norm": 0.11987591534852982, + "learning_rate": 7.941888981527274e-07, + "loss": 0.0092, + "step": 108140 + }, + { + "epoch": 31.466395112016293, + "grad_norm": 0.11374425143003464, + "learning_rate": 7.928622217269421e-07, + "loss": 0.0107, + "step": 108150 + }, + { + "epoch": 31.469304626127435, + "grad_norm": 0.09781233221292496, + "learning_rate": 7.915366242395139e-07, + "loss": 0.0102, + "step": 108160 + }, + { + "epoch": 31.47221414023858, + "grad_norm": 0.17467987537384033, + "learning_rate": 7.902121057911088e-07, + "loss": 0.0117, + "step": 108170 + }, + { + "epoch": 31.475123654349723, + "grad_norm": 0.121156245470047, + "learning_rate": 7.888886664823214e-07, + "loss": 0.0097, + "step": 108180 + }, + { + "epoch": 31.47803316846087, + "grad_norm": 0.13338503241539001, + "learning_rate": 7.875663064136512e-07, + "loss": 0.01, + "step": 108190 + }, + { + "epoch": 31.48094268257201, + "grad_norm": 0.11696972697973251, + "learning_rate": 7.862450256855281e-07, + "loss": 0.01, + "step": 108200 + }, + { + "epoch": 31.483852196683152, + "grad_norm": 0.11656399071216583, + "learning_rate": 7.849248243982915e-07, + "loss": 0.0101, + "step": 108210 + }, + { + "epoch": 31.486761710794298, + "grad_norm": 0.1254146546125412, + "learning_rate": 7.836057026522048e-07, + "loss": 0.012, + "step": 108220 + }, + { + "epoch": 31.48967122490544, + "grad_norm": 0.15225361287593842, + "learning_rate": 7.822876605474427e-07, + "loss": 0.0102, + "step": 108230 + }, + { + "epoch": 31.492580739016585, + "grad_norm": 0.11466509103775024, + "learning_rate": 7.809706981841036e-07, + "loss": 0.0102, + "step": 108240 + }, + { + "epoch": 31.495490253127727, + "grad_norm": 0.08512228727340698, + "learning_rate": 7.796548156622057e-07, + "loss": 0.0107, + "step": 108250 + }, + { + "epoch": 31.498399767238872, + "grad_norm": 0.0906011313199997, + "learning_rate": 7.783400130816759e-07, + "loss": 0.0108, + "step": 108260 + }, + { + "epoch": 31.501309281350014, + "grad_norm": 0.1115039587020874, + "learning_rate": 7.770262905423659e-07, + "loss": 0.0098, + "step": 108270 + }, + { + "epoch": 31.504218795461156, + "grad_norm": 0.1675877571105957, + "learning_rate": 7.757136481440491e-07, + "loss": 0.0098, + "step": 108280 + }, + { + "epoch": 31.507128309572302, + "grad_norm": 0.08302187919616699, + "learning_rate": 7.744020859864092e-07, + "loss": 0.0096, + "step": 108290 + }, + { + "epoch": 31.510037823683444, + "grad_norm": 0.09591837972402573, + "learning_rate": 7.730916041690467e-07, + "loss": 0.0117, + "step": 108300 + }, + { + "epoch": 31.51294733779459, + "grad_norm": 0.13310286402702332, + "learning_rate": 7.717822027914883e-07, + "loss": 0.0129, + "step": 108310 + }, + { + "epoch": 31.51585685190573, + "grad_norm": 0.1580563634634018, + "learning_rate": 7.704738819531764e-07, + "loss": 0.0089, + "step": 108320 + }, + { + "epoch": 31.518766366016877, + "grad_norm": 0.10221477597951889, + "learning_rate": 7.691666417534648e-07, + "loss": 0.0102, + "step": 108330 + }, + { + "epoch": 31.52167588012802, + "grad_norm": 0.18526118993759155, + "learning_rate": 7.678604822916274e-07, + "loss": 0.0101, + "step": 108340 + }, + { + "epoch": 31.52458539423916, + "grad_norm": 0.08451761305332184, + "learning_rate": 7.665554036668682e-07, + "loss": 0.0101, + "step": 108350 + }, + { + "epoch": 31.527494908350306, + "grad_norm": 0.12834417819976807, + "learning_rate": 7.652514059782911e-07, + "loss": 0.011, + "step": 108360 + }, + { + "epoch": 31.530404422461448, + "grad_norm": 0.07597735524177551, + "learning_rate": 7.639484893249288e-07, + "loss": 0.01, + "step": 108370 + }, + { + "epoch": 31.533313936572593, + "grad_norm": 0.09491363912820816, + "learning_rate": 7.626466538057286e-07, + "loss": 0.0102, + "step": 108380 + }, + { + "epoch": 31.536223450683735, + "grad_norm": 0.15173976123332977, + "learning_rate": 7.613458995195582e-07, + "loss": 0.0098, + "step": 108390 + }, + { + "epoch": 31.53913296479488, + "grad_norm": 0.1389288306236267, + "learning_rate": 7.600462265651986e-07, + "loss": 0.0116, + "step": 108400 + }, + { + "epoch": 31.542042478906023, + "grad_norm": 0.10672543942928314, + "learning_rate": 7.587476350413525e-07, + "loss": 0.0111, + "step": 108410 + }, + { + "epoch": 31.544951993017165, + "grad_norm": 0.10481435060501099, + "learning_rate": 7.574501250466376e-07, + "loss": 0.0099, + "step": 108420 + }, + { + "epoch": 31.54786150712831, + "grad_norm": 0.09490357339382172, + "learning_rate": 7.561536966795968e-07, + "loss": 0.0086, + "step": 108430 + }, + { + "epoch": 31.550771021239452, + "grad_norm": 0.0860191285610199, + "learning_rate": 7.548583500386779e-07, + "loss": 0.0099, + "step": 108440 + }, + { + "epoch": 31.553680535350598, + "grad_norm": 0.10451877117156982, + "learning_rate": 7.535640852222591e-07, + "loss": 0.0124, + "step": 108450 + }, + { + "epoch": 31.55659004946174, + "grad_norm": 0.14720484614372253, + "learning_rate": 7.522709023286317e-07, + "loss": 0.0107, + "step": 108460 + }, + { + "epoch": 31.559499563572885, + "grad_norm": 0.0870155617594719, + "learning_rate": 7.509788014559988e-07, + "loss": 0.0113, + "step": 108470 + }, + { + "epoch": 31.562409077684027, + "grad_norm": 0.06232136860489845, + "learning_rate": 7.496877827024934e-07, + "loss": 0.0083, + "step": 108480 + }, + { + "epoch": 31.56531859179517, + "grad_norm": 0.1389746516942978, + "learning_rate": 7.483978461661556e-07, + "loss": 0.0135, + "step": 108490 + }, + { + "epoch": 31.568228105906314, + "grad_norm": 0.1309206336736679, + "learning_rate": 7.471089919449503e-07, + "loss": 0.01, + "step": 108500 + }, + { + "epoch": 31.571137620017456, + "grad_norm": 0.1809508502483368, + "learning_rate": 7.458212201367559e-07, + "loss": 0.0113, + "step": 108510 + }, + { + "epoch": 31.5740471341286, + "grad_norm": 0.10127103328704834, + "learning_rate": 7.445345308393708e-07, + "loss": 0.0103, + "step": 108520 + }, + { + "epoch": 31.576956648239744, + "grad_norm": 0.1478739082813263, + "learning_rate": 7.432489241505086e-07, + "loss": 0.0124, + "step": 108530 + }, + { + "epoch": 31.57986616235089, + "grad_norm": 0.1369992345571518, + "learning_rate": 7.419644001678027e-07, + "loss": 0.0096, + "step": 108540 + }, + { + "epoch": 31.58277567646203, + "grad_norm": 0.0946832075715065, + "learning_rate": 7.406809589888103e-07, + "loss": 0.0114, + "step": 108550 + }, + { + "epoch": 31.585685190573173, + "grad_norm": 0.11892380565404892, + "learning_rate": 7.393986007109915e-07, + "loss": 0.0108, + "step": 108560 + }, + { + "epoch": 31.58859470468432, + "grad_norm": 0.11446026712656021, + "learning_rate": 7.381173254317402e-07, + "loss": 0.0105, + "step": 108570 + }, + { + "epoch": 31.59150421879546, + "grad_norm": 0.13679124414920807, + "learning_rate": 7.36837133248357e-07, + "loss": 0.0102, + "step": 108580 + }, + { + "epoch": 31.594413732906606, + "grad_norm": 0.12235908210277557, + "learning_rate": 7.355580242580656e-07, + "loss": 0.0104, + "step": 108590 + }, + { + "epoch": 31.597323247017748, + "grad_norm": 0.10380154848098755, + "learning_rate": 7.342799985580035e-07, + "loss": 0.0118, + "step": 108600 + }, + { + "epoch": 31.60023276112889, + "grad_norm": 0.13381414115428925, + "learning_rate": 7.330030562452295e-07, + "loss": 0.0109, + "step": 108610 + }, + { + "epoch": 31.603142275240035, + "grad_norm": 0.07168683409690857, + "learning_rate": 7.317271974167211e-07, + "loss": 0.0095, + "step": 108620 + }, + { + "epoch": 31.606051789351177, + "grad_norm": 0.1474916934967041, + "learning_rate": 7.304524221693709e-07, + "loss": 0.0094, + "step": 108630 + }, + { + "epoch": 31.608961303462323, + "grad_norm": 0.08947078138589859, + "learning_rate": 7.29178730599983e-07, + "loss": 0.0102, + "step": 108640 + }, + { + "epoch": 31.611870817573465, + "grad_norm": 0.12340953201055527, + "learning_rate": 7.279061228052952e-07, + "loss": 0.0103, + "step": 108650 + }, + { + "epoch": 31.61478033168461, + "grad_norm": 0.10539325326681137, + "learning_rate": 7.266345988819518e-07, + "loss": 0.011, + "step": 108660 + }, + { + "epoch": 31.617689845795752, + "grad_norm": 0.06238440424203873, + "learning_rate": 7.253641589265103e-07, + "loss": 0.011, + "step": 108670 + }, + { + "epoch": 31.620599359906894, + "grad_norm": 0.11521542817354202, + "learning_rate": 7.240948030354572e-07, + "loss": 0.0081, + "step": 108680 + }, + { + "epoch": 31.62350887401804, + "grad_norm": 0.10123081505298615, + "learning_rate": 7.228265313051935e-07, + "loss": 0.01, + "step": 108690 + }, + { + "epoch": 31.62641838812918, + "grad_norm": 0.10136068612337112, + "learning_rate": 7.215593438320339e-07, + "loss": 0.011, + "step": 108700 + }, + { + "epoch": 31.629327902240327, + "grad_norm": 0.10492564737796783, + "learning_rate": 7.202932407122098e-07, + "loss": 0.0084, + "step": 108710 + }, + { + "epoch": 31.63223741635147, + "grad_norm": 0.12275492399930954, + "learning_rate": 7.190282220418775e-07, + "loss": 0.0101, + "step": 108720 + }, + { + "epoch": 31.635146930462614, + "grad_norm": 0.09442141652107239, + "learning_rate": 7.177642879171053e-07, + "loss": 0.0122, + "step": 108730 + }, + { + "epoch": 31.638056444573756, + "grad_norm": 0.09862438589334488, + "learning_rate": 7.165014384338814e-07, + "loss": 0.0103, + "step": 108740 + }, + { + "epoch": 31.640965958684898, + "grad_norm": 0.10215488076210022, + "learning_rate": 7.152396736881089e-07, + "loss": 0.01, + "step": 108750 + }, + { + "epoch": 31.643875472796044, + "grad_norm": 0.12798267602920532, + "learning_rate": 7.139789937756147e-07, + "loss": 0.0098, + "step": 108760 + }, + { + "epoch": 31.646784986907186, + "grad_norm": 0.10439151525497437, + "learning_rate": 7.127193987921354e-07, + "loss": 0.0115, + "step": 108770 + }, + { + "epoch": 31.64969450101833, + "grad_norm": 0.10075860470533371, + "learning_rate": 7.114608888333279e-07, + "loss": 0.009, + "step": 108780 + }, + { + "epoch": 31.652604015129473, + "grad_norm": 0.11163794994354248, + "learning_rate": 7.102034639947708e-07, + "loss": 0.0109, + "step": 108790 + }, + { + "epoch": 31.65551352924062, + "grad_norm": 0.11293158680200577, + "learning_rate": 7.089471243719575e-07, + "loss": 0.0099, + "step": 108800 + }, + { + "epoch": 31.65842304335176, + "grad_norm": 0.10602504760026932, + "learning_rate": 7.076918700602953e-07, + "loss": 0.0118, + "step": 108810 + }, + { + "epoch": 31.661332557462902, + "grad_norm": 0.29136061668395996, + "learning_rate": 7.06437701155116e-07, + "loss": 0.0099, + "step": 108820 + }, + { + "epoch": 31.664242071574048, + "grad_norm": 0.07644382864236832, + "learning_rate": 7.051846177516635e-07, + "loss": 0.0097, + "step": 108830 + }, + { + "epoch": 31.66715158568519, + "grad_norm": 0.13200487196445465, + "learning_rate": 7.039326199451002e-07, + "loss": 0.0102, + "step": 108840 + }, + { + "epoch": 31.670061099796335, + "grad_norm": 0.10131537169218063, + "learning_rate": 7.026817078305115e-07, + "loss": 0.0093, + "step": 108850 + }, + { + "epoch": 31.672970613907477, + "grad_norm": 0.23050780594348907, + "learning_rate": 7.014318815028914e-07, + "loss": 0.0113, + "step": 108860 + }, + { + "epoch": 31.67588012801862, + "grad_norm": 0.11968255788087845, + "learning_rate": 7.001831410571591e-07, + "loss": 0.0099, + "step": 108870 + }, + { + "epoch": 31.678789642129765, + "grad_norm": 0.07870137691497803, + "learning_rate": 6.989354865881436e-07, + "loss": 0.0108, + "step": 108880 + }, + { + "epoch": 31.681699156240906, + "grad_norm": 0.12956269085407257, + "learning_rate": 6.976889181906027e-07, + "loss": 0.0098, + "step": 108890 + }, + { + "epoch": 31.684608670352052, + "grad_norm": 0.1083388552069664, + "learning_rate": 6.964434359591987e-07, + "loss": 0.0109, + "step": 108900 + }, + { + "epoch": 31.687518184463194, + "grad_norm": 0.1051262840628624, + "learning_rate": 6.95199039988521e-07, + "loss": 0.0102, + "step": 108910 + }, + { + "epoch": 31.69042769857434, + "grad_norm": 0.11298222094774246, + "learning_rate": 6.939557303730743e-07, + "loss": 0.0091, + "step": 108920 + }, + { + "epoch": 31.69333721268548, + "grad_norm": 0.14410999417304993, + "learning_rate": 6.927135072072777e-07, + "loss": 0.0124, + "step": 108930 + }, + { + "epoch": 31.696246726796623, + "grad_norm": 0.09204185009002686, + "learning_rate": 6.914723705854692e-07, + "loss": 0.0098, + "step": 108940 + }, + { + "epoch": 31.69915624090777, + "grad_norm": 0.09973140060901642, + "learning_rate": 6.902323206019051e-07, + "loss": 0.0079, + "step": 108950 + }, + { + "epoch": 31.70206575501891, + "grad_norm": 0.06893222033977509, + "learning_rate": 6.889933573507618e-07, + "loss": 0.0107, + "step": 108960 + }, + { + "epoch": 31.704975269130056, + "grad_norm": 0.08530843257904053, + "learning_rate": 6.877554809261289e-07, + "loss": 0.0102, + "step": 108970 + }, + { + "epoch": 31.707884783241198, + "grad_norm": 0.06420441716909409, + "learning_rate": 6.865186914220128e-07, + "loss": 0.0092, + "step": 108980 + }, + { + "epoch": 31.710794297352344, + "grad_norm": 0.17383909225463867, + "learning_rate": 6.852829889323436e-07, + "loss": 0.0104, + "step": 108990 + }, + { + "epoch": 31.713703811463485, + "grad_norm": 0.23301897943019867, + "learning_rate": 6.840483735509611e-07, + "loss": 0.0111, + "step": 109000 + }, + { + "epoch": 31.716613325574627, + "grad_norm": 0.12896676361560822, + "learning_rate": 6.82814845371627e-07, + "loss": 0.0082, + "step": 109010 + }, + { + "epoch": 31.719522839685773, + "grad_norm": 0.12841925024986267, + "learning_rate": 6.815824044880214e-07, + "loss": 0.0092, + "step": 109020 + }, + { + "epoch": 31.722432353796915, + "grad_norm": 0.13432614505290985, + "learning_rate": 6.803510509937394e-07, + "loss": 0.0117, + "step": 109030 + }, + { + "epoch": 31.72534186790806, + "grad_norm": 0.11475224792957306, + "learning_rate": 6.791207849822945e-07, + "loss": 0.0106, + "step": 109040 + }, + { + "epoch": 31.728251382019202, + "grad_norm": 0.13137900829315186, + "learning_rate": 6.778916065471103e-07, + "loss": 0.0131, + "step": 109050 + }, + { + "epoch": 31.731160896130348, + "grad_norm": 0.12488023936748505, + "learning_rate": 6.76663515781547e-07, + "loss": 0.0091, + "step": 109060 + }, + { + "epoch": 31.73407041024149, + "grad_norm": 0.08162689208984375, + "learning_rate": 6.754365127788636e-07, + "loss": 0.012, + "step": 109070 + }, + { + "epoch": 31.73697992435263, + "grad_norm": 0.13151663541793823, + "learning_rate": 6.742105976322405e-07, + "loss": 0.0096, + "step": 109080 + }, + { + "epoch": 31.739889438463777, + "grad_norm": 0.12818846106529236, + "learning_rate": 6.729857704347797e-07, + "loss": 0.0114, + "step": 109090 + }, + { + "epoch": 31.74279895257492, + "grad_norm": 0.09523124992847443, + "learning_rate": 6.717620312795036e-07, + "loss": 0.0099, + "step": 109100 + }, + { + "epoch": 31.745708466686064, + "grad_norm": 0.11550548672676086, + "learning_rate": 6.705393802593396e-07, + "loss": 0.01, + "step": 109110 + }, + { + "epoch": 31.748617980797206, + "grad_norm": 0.1213279515504837, + "learning_rate": 6.693178174671466e-07, + "loss": 0.0104, + "step": 109120 + }, + { + "epoch": 31.751527494908352, + "grad_norm": 0.07776990532875061, + "learning_rate": 6.680973429956888e-07, + "loss": 0.0099, + "step": 109130 + }, + { + "epoch": 31.754437009019494, + "grad_norm": 0.0968211218714714, + "learning_rate": 6.668779569376571e-07, + "loss": 0.0106, + "step": 109140 + }, + { + "epoch": 31.757346523130636, + "grad_norm": 0.08437841385602951, + "learning_rate": 6.656596593856523e-07, + "loss": 0.0093, + "step": 109150 + }, + { + "epoch": 31.76025603724178, + "grad_norm": 0.09584114700555801, + "learning_rate": 6.644424504321972e-07, + "loss": 0.0109, + "step": 109160 + }, + { + "epoch": 31.763165551352923, + "grad_norm": 0.10183250159025192, + "learning_rate": 6.632263301697345e-07, + "loss": 0.0108, + "step": 109170 + }, + { + "epoch": 31.76607506546407, + "grad_norm": 0.13275212049484253, + "learning_rate": 6.620112986906152e-07, + "loss": 0.0122, + "step": 109180 + }, + { + "epoch": 31.76898457957521, + "grad_norm": 0.17219239473342896, + "learning_rate": 6.607973560871173e-07, + "loss": 0.0099, + "step": 109190 + }, + { + "epoch": 31.771894093686356, + "grad_norm": 0.08981399238109589, + "learning_rate": 6.595845024514285e-07, + "loss": 0.0116, + "step": 109200 + }, + { + "epoch": 31.774803607797498, + "grad_norm": 0.07844484597444534, + "learning_rate": 6.583727378756571e-07, + "loss": 0.0102, + "step": 109210 + }, + { + "epoch": 31.77771312190864, + "grad_norm": 0.08904692530632019, + "learning_rate": 6.571620624518326e-07, + "loss": 0.0128, + "step": 109220 + }, + { + "epoch": 31.780622636019785, + "grad_norm": 0.17677344381809235, + "learning_rate": 6.559524762718933e-07, + "loss": 0.0106, + "step": 109230 + }, + { + "epoch": 31.783532150130927, + "grad_norm": 0.13520687818527222, + "learning_rate": 6.547439794277005e-07, + "loss": 0.0095, + "step": 109240 + }, + { + "epoch": 31.786441664242073, + "grad_norm": 0.06889946758747101, + "learning_rate": 6.535365720110326e-07, + "loss": 0.009, + "step": 109250 + }, + { + "epoch": 31.789351178353215, + "grad_norm": 0.07992031425237656, + "learning_rate": 6.523302541135845e-07, + "loss": 0.0093, + "step": 109260 + }, + { + "epoch": 31.79226069246436, + "grad_norm": 0.0660422295331955, + "learning_rate": 6.511250258269647e-07, + "loss": 0.0079, + "step": 109270 + }, + { + "epoch": 31.795170206575502, + "grad_norm": 0.1658865213394165, + "learning_rate": 6.499208872427065e-07, + "loss": 0.01, + "step": 109280 + }, + { + "epoch": 31.798079720686644, + "grad_norm": 0.07787895202636719, + "learning_rate": 6.487178384522552e-07, + "loss": 0.0088, + "step": 109290 + }, + { + "epoch": 31.80098923479779, + "grad_norm": 0.131865993142128, + "learning_rate": 6.475158795469743e-07, + "loss": 0.0091, + "step": 109300 + }, + { + "epoch": 31.80389874890893, + "grad_norm": 0.1301511824131012, + "learning_rate": 6.463150106181426e-07, + "loss": 0.0102, + "step": 109310 + }, + { + "epoch": 31.806808263020077, + "grad_norm": 0.18196681141853333, + "learning_rate": 6.451152317569587e-07, + "loss": 0.0121, + "step": 109320 + }, + { + "epoch": 31.80971777713122, + "grad_norm": 0.08443278819322586, + "learning_rate": 6.439165430545413e-07, + "loss": 0.011, + "step": 109330 + }, + { + "epoch": 31.81262729124236, + "grad_norm": 0.09034670889377594, + "learning_rate": 6.42718944601921e-07, + "loss": 0.0099, + "step": 109340 + }, + { + "epoch": 31.815536805353506, + "grad_norm": 0.10562459379434586, + "learning_rate": 6.415224364900435e-07, + "loss": 0.0112, + "step": 109350 + }, + { + "epoch": 31.81844631946465, + "grad_norm": 0.15094095468521118, + "learning_rate": 6.403270188097827e-07, + "loss": 0.0114, + "step": 109360 + }, + { + "epoch": 31.821355833575794, + "grad_norm": 0.14621177315711975, + "learning_rate": 6.391326916519192e-07, + "loss": 0.0109, + "step": 109370 + }, + { + "epoch": 31.824265347686936, + "grad_norm": 0.09624097496271133, + "learning_rate": 6.379394551071521e-07, + "loss": 0.009, + "step": 109380 + }, + { + "epoch": 31.82717486179808, + "grad_norm": 0.07902518659830093, + "learning_rate": 6.367473092661025e-07, + "loss": 0.0117, + "step": 109390 + }, + { + "epoch": 31.830084375909223, + "grad_norm": 0.12327911704778671, + "learning_rate": 6.355562542193078e-07, + "loss": 0.0094, + "step": 109400 + }, + { + "epoch": 31.832993890020365, + "grad_norm": 0.08453162759542465, + "learning_rate": 6.343662900572172e-07, + "loss": 0.0108, + "step": 109410 + }, + { + "epoch": 31.83590340413151, + "grad_norm": 0.08094175159931183, + "learning_rate": 6.33177416870202e-07, + "loss": 0.0087, + "step": 109420 + }, + { + "epoch": 31.838812918242652, + "grad_norm": 0.15256205201148987, + "learning_rate": 6.319896347485466e-07, + "loss": 0.012, + "step": 109430 + }, + { + "epoch": 31.841722432353798, + "grad_norm": 0.11254730820655823, + "learning_rate": 6.30802943782462e-07, + "loss": 0.01, + "step": 109440 + }, + { + "epoch": 31.84463194646494, + "grad_norm": 0.09835207462310791, + "learning_rate": 6.296173440620628e-07, + "loss": 0.0095, + "step": 109450 + }, + { + "epoch": 31.847541460576085, + "grad_norm": 0.13734042644500732, + "learning_rate": 6.284328356773889e-07, + "loss": 0.0092, + "step": 109460 + }, + { + "epoch": 31.850450974687227, + "grad_norm": 0.08934329450130463, + "learning_rate": 6.272494187183997e-07, + "loss": 0.0087, + "step": 109470 + }, + { + "epoch": 31.85336048879837, + "grad_norm": 0.10653985291719437, + "learning_rate": 6.260670932749635e-07, + "loss": 0.0102, + "step": 109480 + }, + { + "epoch": 31.856270002909515, + "grad_norm": 0.15239480137825012, + "learning_rate": 6.24885859436875e-07, + "loss": 0.0104, + "step": 109490 + }, + { + "epoch": 31.859179517020657, + "grad_norm": 0.09840337187051773, + "learning_rate": 6.237057172938343e-07, + "loss": 0.0096, + "step": 109500 + }, + { + "epoch": 31.862089031131802, + "grad_norm": 0.15969878435134888, + "learning_rate": 6.225266669354713e-07, + "loss": 0.0099, + "step": 109510 + }, + { + "epoch": 31.864998545242944, + "grad_norm": 0.09532878547906876, + "learning_rate": 6.213487084513242e-07, + "loss": 0.0113, + "step": 109520 + }, + { + "epoch": 31.86790805935409, + "grad_norm": 0.14271077513694763, + "learning_rate": 6.201718419308517e-07, + "loss": 0.0086, + "step": 109530 + }, + { + "epoch": 31.87081757346523, + "grad_norm": 0.14006371796131134, + "learning_rate": 6.189960674634287e-07, + "loss": 0.0107, + "step": 109540 + }, + { + "epoch": 31.873727087576373, + "grad_norm": 0.11789991706609726, + "learning_rate": 6.178213851383474e-07, + "loss": 0.01, + "step": 109550 + }, + { + "epoch": 31.87663660168752, + "grad_norm": 0.1037549152970314, + "learning_rate": 6.166477950448213e-07, + "loss": 0.0083, + "step": 109560 + }, + { + "epoch": 31.87954611579866, + "grad_norm": 0.1076866015791893, + "learning_rate": 6.154752972719691e-07, + "loss": 0.0105, + "step": 109570 + }, + { + "epoch": 31.882455629909806, + "grad_norm": 0.14943504333496094, + "learning_rate": 6.143038919088395e-07, + "loss": 0.0095, + "step": 109580 + }, + { + "epoch": 31.88536514402095, + "grad_norm": 0.09928181022405624, + "learning_rate": 6.131335790443932e-07, + "loss": 0.0095, + "step": 109590 + }, + { + "epoch": 31.88827465813209, + "grad_norm": 0.14133302867412567, + "learning_rate": 6.119643587675073e-07, + "loss": 0.011, + "step": 109600 + }, + { + "epoch": 31.891184172243236, + "grad_norm": 0.07721342146396637, + "learning_rate": 6.10796231166974e-07, + "loss": 0.0095, + "step": 109610 + }, + { + "epoch": 31.894093686354378, + "grad_norm": 0.1597939282655716, + "learning_rate": 6.096291963315059e-07, + "loss": 0.0114, + "step": 109620 + }, + { + "epoch": 31.897003200465523, + "grad_norm": 0.12081795185804367, + "learning_rate": 6.084632543497337e-07, + "loss": 0.01, + "step": 109630 + }, + { + "epoch": 31.899912714576665, + "grad_norm": 0.09710420668125153, + "learning_rate": 6.072984053102032e-07, + "loss": 0.0101, + "step": 109640 + }, + { + "epoch": 31.90282222868781, + "grad_norm": 0.07039511948823929, + "learning_rate": 6.061346493013686e-07, + "loss": 0.0102, + "step": 109650 + }, + { + "epoch": 31.905731742798952, + "grad_norm": 0.12312440574169159, + "learning_rate": 6.049719864116227e-07, + "loss": 0.0131, + "step": 109660 + }, + { + "epoch": 31.908641256910094, + "grad_norm": 0.08751139044761658, + "learning_rate": 6.038104167292546e-07, + "loss": 0.0105, + "step": 109670 + }, + { + "epoch": 31.91155077102124, + "grad_norm": 0.125532865524292, + "learning_rate": 6.026499403424757e-07, + "loss": 0.0118, + "step": 109680 + }, + { + "epoch": 31.91446028513238, + "grad_norm": 0.09237536042928696, + "learning_rate": 6.014905573394202e-07, + "loss": 0.0107, + "step": 109690 + }, + { + "epoch": 31.917369799243527, + "grad_norm": 0.08037266880273819, + "learning_rate": 6.00332267808138e-07, + "loss": 0.0107, + "step": 109700 + }, + { + "epoch": 31.92027931335467, + "grad_norm": 0.13146348297595978, + "learning_rate": 5.991750718365885e-07, + "loss": 0.0099, + "step": 109710 + }, + { + "epoch": 31.923188827465815, + "grad_norm": 0.10408502072095871, + "learning_rate": 5.980189695126547e-07, + "loss": 0.0093, + "step": 109720 + }, + { + "epoch": 31.926098341576957, + "grad_norm": 0.10190797597169876, + "learning_rate": 5.968639609241349e-07, + "loss": 0.0105, + "step": 109730 + }, + { + "epoch": 31.9290078556881, + "grad_norm": 0.12258736789226532, + "learning_rate": 5.957100461587472e-07, + "loss": 0.0107, + "step": 109740 + }, + { + "epoch": 31.931917369799244, + "grad_norm": 0.09874217212200165, + "learning_rate": 5.945572253041198e-07, + "loss": 0.0089, + "step": 109750 + }, + { + "epoch": 31.934826883910386, + "grad_norm": 0.1533600389957428, + "learning_rate": 5.934054984478026e-07, + "loss": 0.0097, + "step": 109760 + }, + { + "epoch": 31.93773639802153, + "grad_norm": 0.09090138226747513, + "learning_rate": 5.922548656772641e-07, + "loss": 0.01, + "step": 109770 + }, + { + "epoch": 31.940645912132673, + "grad_norm": 0.11029686778783798, + "learning_rate": 5.911053270798861e-07, + "loss": 0.0106, + "step": 109780 + }, + { + "epoch": 31.94355542624382, + "grad_norm": 0.1376999318599701, + "learning_rate": 5.899568827429668e-07, + "loss": 0.0091, + "step": 109790 + }, + { + "epoch": 31.94646494035496, + "grad_norm": 0.11007507145404816, + "learning_rate": 5.888095327537235e-07, + "loss": 0.0094, + "step": 109800 + }, + { + "epoch": 31.949374454466103, + "grad_norm": 0.10399533808231354, + "learning_rate": 5.876632771992929e-07, + "loss": 0.0095, + "step": 109810 + }, + { + "epoch": 31.952283968577248, + "grad_norm": 0.0990951731801033, + "learning_rate": 5.865181161667221e-07, + "loss": 0.0092, + "step": 109820 + }, + { + "epoch": 31.95519348268839, + "grad_norm": 0.1413932591676712, + "learning_rate": 5.8537404974298e-07, + "loss": 0.0101, + "step": 109830 + }, + { + "epoch": 31.958102996799536, + "grad_norm": 0.11528007686138153, + "learning_rate": 5.842310780149502e-07, + "loss": 0.0111, + "step": 109840 + }, + { + "epoch": 31.961012510910678, + "grad_norm": 0.07328586280345917, + "learning_rate": 5.830892010694333e-07, + "loss": 0.0124, + "step": 109850 + }, + { + "epoch": 31.963922025021823, + "grad_norm": 0.096938356757164, + "learning_rate": 5.819484189931501e-07, + "loss": 0.0097, + "step": 109860 + }, + { + "epoch": 31.966831539132965, + "grad_norm": 0.13627147674560547, + "learning_rate": 5.808087318727329e-07, + "loss": 0.0102, + "step": 109870 + }, + { + "epoch": 31.969741053244107, + "grad_norm": 0.10983692854642868, + "learning_rate": 5.796701397947341e-07, + "loss": 0.0106, + "step": 109880 + }, + { + "epoch": 31.972650567355252, + "grad_norm": 0.06913948804140091, + "learning_rate": 5.785326428456211e-07, + "loss": 0.0086, + "step": 109890 + }, + { + "epoch": 31.975560081466394, + "grad_norm": 0.10427329689264297, + "learning_rate": 5.773962411117834e-07, + "loss": 0.0109, + "step": 109900 + }, + { + "epoch": 31.97846959557754, + "grad_norm": 0.09712781757116318, + "learning_rate": 5.762609346795184e-07, + "loss": 0.0143, + "step": 109910 + }, + { + "epoch": 31.98137910968868, + "grad_norm": 0.09159762412309647, + "learning_rate": 5.751267236350454e-07, + "loss": 0.0108, + "step": 109920 + }, + { + "epoch": 31.984288623799827, + "grad_norm": 0.10268209874629974, + "learning_rate": 5.739936080645041e-07, + "loss": 0.0103, + "step": 109930 + }, + { + "epoch": 31.98719813791097, + "grad_norm": 0.0755164697766304, + "learning_rate": 5.728615880539456e-07, + "loss": 0.0103, + "step": 109940 + }, + { + "epoch": 31.99010765202211, + "grad_norm": 0.0887598842382431, + "learning_rate": 5.717306636893344e-07, + "loss": 0.0104, + "step": 109950 + }, + { + "epoch": 31.993017166133257, + "grad_norm": 0.0940188392996788, + "learning_rate": 5.70600835056565e-07, + "loss": 0.0096, + "step": 109960 + }, + { + "epoch": 31.9959266802444, + "grad_norm": 0.3889932632446289, + "learning_rate": 5.694721022414373e-07, + "loss": 0.0113, + "step": 109970 + }, + { + "epoch": 31.998836194355544, + "grad_norm": 0.07307486236095428, + "learning_rate": 5.683444653296676e-07, + "loss": 0.0075, + "step": 109980 + }, + { + "epoch": 32.001745708466686, + "grad_norm": 0.1480894535779953, + "learning_rate": 5.672179244068959e-07, + "loss": 0.0089, + "step": 109990 + }, + { + "epoch": 32.00465522257783, + "grad_norm": 0.09260493516921997, + "learning_rate": 5.660924795586753e-07, + "loss": 0.0126, + "step": 110000 + }, + { + "epoch": 32.00756473668897, + "grad_norm": 0.10900694876909256, + "learning_rate": 5.649681308704758e-07, + "loss": 0.0095, + "step": 110010 + }, + { + "epoch": 32.010474250800115, + "grad_norm": 0.3491433262825012, + "learning_rate": 5.638448784276822e-07, + "loss": 0.01, + "step": 110020 + }, + { + "epoch": 32.01338376491126, + "grad_norm": 0.12995365262031555, + "learning_rate": 5.627227223156001e-07, + "loss": 0.0116, + "step": 110030 + }, + { + "epoch": 32.016293279022406, + "grad_norm": 0.0968073382973671, + "learning_rate": 5.616016626194492e-07, + "loss": 0.0103, + "step": 110040 + }, + { + "epoch": 32.019202793133545, + "grad_norm": 0.11376890540122986, + "learning_rate": 5.604816994243683e-07, + "loss": 0.0093, + "step": 110050 + }, + { + "epoch": 32.02211230724469, + "grad_norm": 0.08295627683401108, + "learning_rate": 5.593628328154077e-07, + "loss": 0.0088, + "step": 110060 + }, + { + "epoch": 32.025021821355836, + "grad_norm": 0.09366970509290695, + "learning_rate": 5.582450628775426e-07, + "loss": 0.0107, + "step": 110070 + }, + { + "epoch": 32.027931335466974, + "grad_norm": 0.16708675026893616, + "learning_rate": 5.571283896956569e-07, + "loss": 0.0107, + "step": 110080 + }, + { + "epoch": 32.03084084957812, + "grad_norm": 0.09746364504098892, + "learning_rate": 5.560128133545545e-07, + "loss": 0.0096, + "step": 110090 + }, + { + "epoch": 32.033750363689265, + "grad_norm": 0.09079255908727646, + "learning_rate": 5.548983339389557e-07, + "loss": 0.0101, + "step": 110100 + }, + { + "epoch": 32.03665987780041, + "grad_norm": 0.11083856970071793, + "learning_rate": 5.537849515335014e-07, + "loss": 0.0107, + "step": 110110 + }, + { + "epoch": 32.03956939191155, + "grad_norm": 0.07990050315856934, + "learning_rate": 5.526726662227421e-07, + "loss": 0.0104, + "step": 110120 + }, + { + "epoch": 32.042478906022694, + "grad_norm": 0.09091328084468842, + "learning_rate": 5.515614780911521e-07, + "loss": 0.0104, + "step": 110130 + }, + { + "epoch": 32.04538842013384, + "grad_norm": 0.09374379366636276, + "learning_rate": 5.504513872231137e-07, + "loss": 0.0113, + "step": 110140 + }, + { + "epoch": 32.04829793424498, + "grad_norm": 0.15270808339118958, + "learning_rate": 5.493423937029346e-07, + "loss": 0.0096, + "step": 110150 + }, + { + "epoch": 32.051207448356124, + "grad_norm": 0.13539308309555054, + "learning_rate": 5.48234497614834e-07, + "loss": 0.012, + "step": 110160 + }, + { + "epoch": 32.05411696246727, + "grad_norm": 0.10375829041004181, + "learning_rate": 5.471276990429497e-07, + "loss": 0.01, + "step": 110170 + }, + { + "epoch": 32.057026476578415, + "grad_norm": 0.1230444461107254, + "learning_rate": 5.460219980713377e-07, + "loss": 0.011, + "step": 110180 + }, + { + "epoch": 32.05993599068955, + "grad_norm": 0.11785105615854263, + "learning_rate": 5.449173947839659e-07, + "loss": 0.0099, + "step": 110190 + }, + { + "epoch": 32.0628455048007, + "grad_norm": 0.10912847518920898, + "learning_rate": 5.438138892647237e-07, + "loss": 0.0108, + "step": 110200 + }, + { + "epoch": 32.065755018911844, + "grad_norm": 0.1549524962902069, + "learning_rate": 5.427114815974127e-07, + "loss": 0.0097, + "step": 110210 + }, + { + "epoch": 32.06866453302298, + "grad_norm": 0.14858147501945496, + "learning_rate": 5.416101718657556e-07, + "loss": 0.0104, + "step": 110220 + }, + { + "epoch": 32.07157404713413, + "grad_norm": 0.07729260623455048, + "learning_rate": 5.405099601533892e-07, + "loss": 0.0117, + "step": 110230 + }, + { + "epoch": 32.07448356124527, + "grad_norm": 0.11935745924711227, + "learning_rate": 5.394108465438679e-07, + "loss": 0.0102, + "step": 110240 + }, + { + "epoch": 32.07739307535642, + "grad_norm": 0.10141675174236298, + "learning_rate": 5.383128311206586e-07, + "loss": 0.0109, + "step": 110250 + }, + { + "epoch": 32.08030258946756, + "grad_norm": 0.1124369353055954, + "learning_rate": 5.372159139671512e-07, + "loss": 0.0103, + "step": 110260 + }, + { + "epoch": 32.0832121035787, + "grad_norm": 0.1011970043182373, + "learning_rate": 5.361200951666522e-07, + "loss": 0.0087, + "step": 110270 + }, + { + "epoch": 32.08612161768985, + "grad_norm": 0.08529413491487503, + "learning_rate": 5.350253748023754e-07, + "loss": 0.0105, + "step": 110280 + }, + { + "epoch": 32.089031131800986, + "grad_norm": 0.09971300512552261, + "learning_rate": 5.339317529574606e-07, + "loss": 0.0093, + "step": 110290 + }, + { + "epoch": 32.09194064591213, + "grad_norm": 0.07479105144739151, + "learning_rate": 5.328392297149647e-07, + "loss": 0.0099, + "step": 110300 + }, + { + "epoch": 32.09485016002328, + "grad_norm": 0.11990593373775482, + "learning_rate": 5.317478051578534e-07, + "loss": 0.0105, + "step": 110310 + }, + { + "epoch": 32.09775967413442, + "grad_norm": 0.08928803354501724, + "learning_rate": 5.306574793690117e-07, + "loss": 0.0111, + "step": 110320 + }, + { + "epoch": 32.10066918824556, + "grad_norm": 0.08716239780187607, + "learning_rate": 5.295682524312467e-07, + "loss": 0.0105, + "step": 110330 + }, + { + "epoch": 32.10357870235671, + "grad_norm": 0.07820849120616913, + "learning_rate": 5.284801244272774e-07, + "loss": 0.0099, + "step": 110340 + }, + { + "epoch": 32.10648821646785, + "grad_norm": 0.08609037101268768, + "learning_rate": 5.273930954397393e-07, + "loss": 0.01, + "step": 110350 + }, + { + "epoch": 32.10939773057899, + "grad_norm": 0.08315495401620865, + "learning_rate": 5.263071655511798e-07, + "loss": 0.0099, + "step": 110360 + }, + { + "epoch": 32.112307244690136, + "grad_norm": 0.11264033615589142, + "learning_rate": 5.252223348440776e-07, + "loss": 0.0105, + "step": 110370 + }, + { + "epoch": 32.11521675880128, + "grad_norm": 0.07476692646741867, + "learning_rate": 5.241386034008139e-07, + "loss": 0.0085, + "step": 110380 + }, + { + "epoch": 32.11812627291243, + "grad_norm": 0.11191030591726303, + "learning_rate": 5.230559713036892e-07, + "loss": 0.0103, + "step": 110390 + }, + { + "epoch": 32.121035787023565, + "grad_norm": 0.1089220717549324, + "learning_rate": 5.219744386349229e-07, + "loss": 0.0123, + "step": 110400 + }, + { + "epoch": 32.12394530113471, + "grad_norm": 0.1088428869843483, + "learning_rate": 5.208940054766543e-07, + "loss": 0.0084, + "step": 110410 + }, + { + "epoch": 32.126854815245856, + "grad_norm": 0.07637573778629303, + "learning_rate": 5.198146719109309e-07, + "loss": 0.0111, + "step": 110420 + }, + { + "epoch": 32.129764329356995, + "grad_norm": 0.07246862351894379, + "learning_rate": 5.187364380197224e-07, + "loss": 0.0113, + "step": 110430 + }, + { + "epoch": 32.13267384346814, + "grad_norm": 0.11853985488414764, + "learning_rate": 5.176593038849131e-07, + "loss": 0.0108, + "step": 110440 + }, + { + "epoch": 32.135583357579286, + "grad_norm": 0.1336696743965149, + "learning_rate": 5.165832695883043e-07, + "loss": 0.0129, + "step": 110450 + }, + { + "epoch": 32.13849287169043, + "grad_norm": 0.10671409219503403, + "learning_rate": 5.155083352116141e-07, + "loss": 0.0102, + "step": 110460 + }, + { + "epoch": 32.14140238580157, + "grad_norm": 0.11619283258914948, + "learning_rate": 5.144345008364753e-07, + "loss": 0.0112, + "step": 110470 + }, + { + "epoch": 32.144311899912715, + "grad_norm": 0.10700175911188126, + "learning_rate": 5.133617665444429e-07, + "loss": 0.0107, + "step": 110480 + }, + { + "epoch": 32.14722141402386, + "grad_norm": 0.10344427078962326, + "learning_rate": 5.122901324169782e-07, + "loss": 0.0098, + "step": 110490 + }, + { + "epoch": 32.150130928135, + "grad_norm": 0.11591722071170807, + "learning_rate": 5.112195985354695e-07, + "loss": 0.0101, + "step": 110500 + }, + { + "epoch": 32.153040442246144, + "grad_norm": 0.05743898078799248, + "learning_rate": 5.101501649812118e-07, + "loss": 0.009, + "step": 110510 + }, + { + "epoch": 32.15594995635729, + "grad_norm": 0.08118270337581635, + "learning_rate": 5.090818318354285e-07, + "loss": 0.0103, + "step": 110520 + }, + { + "epoch": 32.15885947046843, + "grad_norm": 0.12732477486133575, + "learning_rate": 5.080145991792445e-07, + "loss": 0.0132, + "step": 110530 + }, + { + "epoch": 32.161768984579574, + "grad_norm": 0.08107178658246994, + "learning_rate": 5.069484670937152e-07, + "loss": 0.0105, + "step": 110540 + }, + { + "epoch": 32.16467849869072, + "grad_norm": 0.10749290138483047, + "learning_rate": 5.058834356598024e-07, + "loss": 0.0107, + "step": 110550 + }, + { + "epoch": 32.167588012801865, + "grad_norm": 0.09776169061660767, + "learning_rate": 5.048195049583898e-07, + "loss": 0.01, + "step": 110560 + }, + { + "epoch": 32.170497526913, + "grad_norm": 0.10464395582675934, + "learning_rate": 5.037566750702777e-07, + "loss": 0.0104, + "step": 110570 + }, + { + "epoch": 32.17340704102415, + "grad_norm": 0.10297301411628723, + "learning_rate": 5.026949460761765e-07, + "loss": 0.0098, + "step": 110580 + }, + { + "epoch": 32.176316555135294, + "grad_norm": 0.1304662525653839, + "learning_rate": 5.016343180567217e-07, + "loss": 0.0109, + "step": 110590 + }, + { + "epoch": 32.17922606924643, + "grad_norm": 0.07984668016433716, + "learning_rate": 5.005747910924607e-07, + "loss": 0.0087, + "step": 110600 + }, + { + "epoch": 32.18213558335758, + "grad_norm": 0.07351759076118469, + "learning_rate": 4.995163652638556e-07, + "loss": 0.0089, + "step": 110610 + }, + { + "epoch": 32.18504509746872, + "grad_norm": 0.06648815423250198, + "learning_rate": 4.984590406512874e-07, + "loss": 0.0089, + "step": 110620 + }, + { + "epoch": 32.18795461157987, + "grad_norm": 0.1430829018354416, + "learning_rate": 4.974028173350531e-07, + "loss": 0.0117, + "step": 110630 + }, + { + "epoch": 32.19086412569101, + "grad_norm": 0.11194227635860443, + "learning_rate": 4.963476953953672e-07, + "loss": 0.009, + "step": 110640 + }, + { + "epoch": 32.19377363980215, + "grad_norm": 0.10054033249616623, + "learning_rate": 4.952936749123572e-07, + "loss": 0.0102, + "step": 110650 + }, + { + "epoch": 32.1966831539133, + "grad_norm": 0.10551789402961731, + "learning_rate": 4.942407559660689e-07, + "loss": 0.0106, + "step": 110660 + }, + { + "epoch": 32.19959266802444, + "grad_norm": 0.08278116583824158, + "learning_rate": 4.931889386364669e-07, + "loss": 0.0097, + "step": 110670 + }, + { + "epoch": 32.20250218213558, + "grad_norm": 0.12361503392457962, + "learning_rate": 4.921382230034288e-07, + "loss": 0.0111, + "step": 110680 + }, + { + "epoch": 32.20541169624673, + "grad_norm": 0.106845922768116, + "learning_rate": 4.910886091467476e-07, + "loss": 0.0113, + "step": 110690 + }, + { + "epoch": 32.20832121035787, + "grad_norm": 0.19084632396697998, + "learning_rate": 4.900400971461361e-07, + "loss": 0.0112, + "step": 110700 + }, + { + "epoch": 32.21123072446901, + "grad_norm": 0.13887165486812592, + "learning_rate": 4.889926870812222e-07, + "loss": 0.0104, + "step": 110710 + }, + { + "epoch": 32.21414023858016, + "grad_norm": 0.08551866561174393, + "learning_rate": 4.879463790315508e-07, + "loss": 0.0114, + "step": 110720 + }, + { + "epoch": 32.2170497526913, + "grad_norm": 0.07743681967258453, + "learning_rate": 4.869011730765799e-07, + "loss": 0.0097, + "step": 110730 + }, + { + "epoch": 32.21995926680244, + "grad_norm": 0.09723922610282898, + "learning_rate": 4.858570692956843e-07, + "loss": 0.0093, + "step": 110740 + }, + { + "epoch": 32.222868780913586, + "grad_norm": 0.12477528303861618, + "learning_rate": 4.848140677681623e-07, + "loss": 0.0103, + "step": 110750 + }, + { + "epoch": 32.22577829502473, + "grad_norm": 0.09990587830543518, + "learning_rate": 4.837721685732188e-07, + "loss": 0.0093, + "step": 110760 + }, + { + "epoch": 32.22868780913588, + "grad_norm": 0.06492569297552109, + "learning_rate": 4.827313717899807e-07, + "loss": 0.0117, + "step": 110770 + }, + { + "epoch": 32.231597323247016, + "grad_norm": 0.11801129579544067, + "learning_rate": 4.816916774974894e-07, + "loss": 0.0098, + "step": 110780 + }, + { + "epoch": 32.23450683735816, + "grad_norm": 0.09648147225379944, + "learning_rate": 4.806530857747021e-07, + "loss": 0.0128, + "step": 110790 + }, + { + "epoch": 32.23741635146931, + "grad_norm": 0.10513093322515488, + "learning_rate": 4.796155967004939e-07, + "loss": 0.012, + "step": 110800 + }, + { + "epoch": 32.240325865580445, + "grad_norm": 0.09351974725723267, + "learning_rate": 4.785792103536551e-07, + "loss": 0.012, + "step": 110810 + }, + { + "epoch": 32.24323537969159, + "grad_norm": 0.10203740000724792, + "learning_rate": 4.775439268128928e-07, + "loss": 0.0109, + "step": 110820 + }, + { + "epoch": 32.246144893802736, + "grad_norm": 0.08104202151298523, + "learning_rate": 4.765097461568274e-07, + "loss": 0.0114, + "step": 110830 + }, + { + "epoch": 32.24905440791388, + "grad_norm": 0.09138374775648117, + "learning_rate": 4.7547666846400275e-07, + "loss": 0.0115, + "step": 110840 + }, + { + "epoch": 32.25196392202502, + "grad_norm": 0.09559999406337738, + "learning_rate": 4.744446938128694e-07, + "loss": 0.011, + "step": 110850 + }, + { + "epoch": 32.254873436136165, + "grad_norm": 0.12647870182991028, + "learning_rate": 4.734138222818013e-07, + "loss": 0.0089, + "step": 110860 + }, + { + "epoch": 32.25778295024731, + "grad_norm": 0.08500046283006668, + "learning_rate": 4.723840539490876e-07, + "loss": 0.012, + "step": 110870 + }, + { + "epoch": 32.26069246435845, + "grad_norm": 0.11937836557626724, + "learning_rate": 4.713553888929306e-07, + "loss": 0.0099, + "step": 110880 + }, + { + "epoch": 32.263601978469595, + "grad_norm": 0.11506782472133636, + "learning_rate": 4.703278271914496e-07, + "loss": 0.0112, + "step": 110890 + }, + { + "epoch": 32.26651149258074, + "grad_norm": 0.08518647402524948, + "learning_rate": 4.6930136892268547e-07, + "loss": 0.0104, + "step": 110900 + }, + { + "epoch": 32.269421006691886, + "grad_norm": 0.19467616081237793, + "learning_rate": 4.6827601416458756e-07, + "loss": 0.0105, + "step": 110910 + }, + { + "epoch": 32.272330520803024, + "grad_norm": 0.07783202081918716, + "learning_rate": 4.6725176299502346e-07, + "loss": 0.008, + "step": 110920 + }, + { + "epoch": 32.27524003491417, + "grad_norm": 0.1379561573266983, + "learning_rate": 4.662286154917794e-07, + "loss": 0.0092, + "step": 110930 + }, + { + "epoch": 32.278149549025315, + "grad_norm": 0.11104656010866165, + "learning_rate": 4.6520657173256155e-07, + "loss": 0.0111, + "step": 110940 + }, + { + "epoch": 32.28105906313645, + "grad_norm": 0.0926966443657875, + "learning_rate": 4.641856317949811e-07, + "loss": 0.0088, + "step": 110950 + }, + { + "epoch": 32.2839685772476, + "grad_norm": 0.08550000935792923, + "learning_rate": 4.63165795756571e-07, + "loss": 0.012, + "step": 110960 + }, + { + "epoch": 32.286878091358744, + "grad_norm": 0.12177462875843048, + "learning_rate": 4.6214706369478775e-07, + "loss": 0.0089, + "step": 110970 + }, + { + "epoch": 32.28978760546989, + "grad_norm": 0.10299546271562576, + "learning_rate": 4.611294356869944e-07, + "loss": 0.01, + "step": 110980 + }, + { + "epoch": 32.29269711958103, + "grad_norm": 0.12372317910194397, + "learning_rate": 4.601129118104691e-07, + "loss": 0.0101, + "step": 110990 + }, + { + "epoch": 32.295606633692174, + "grad_norm": 0.09771998971700668, + "learning_rate": 4.590974921424135e-07, + "loss": 0.0105, + "step": 111000 + }, + { + "epoch": 32.29851614780332, + "grad_norm": 0.16848602890968323, + "learning_rate": 4.5808317675994416e-07, + "loss": 0.0103, + "step": 111010 + }, + { + "epoch": 32.30142566191446, + "grad_norm": 0.07659286260604858, + "learning_rate": 4.5706996574008787e-07, + "loss": 0.009, + "step": 111020 + }, + { + "epoch": 32.3043351760256, + "grad_norm": 0.11169321089982986, + "learning_rate": 4.560578591597914e-07, + "loss": 0.0128, + "step": 111030 + }, + { + "epoch": 32.30724469013675, + "grad_norm": 0.1049814373254776, + "learning_rate": 4.550468570959199e-07, + "loss": 0.0117, + "step": 111040 + }, + { + "epoch": 32.310154204247894, + "grad_norm": 0.10528922826051712, + "learning_rate": 4.540369596252519e-07, + "loss": 0.0092, + "step": 111050 + }, + { + "epoch": 32.31306371835903, + "grad_norm": 0.10050978511571884, + "learning_rate": 4.530281668244812e-07, + "loss": 0.0111, + "step": 111060 + }, + { + "epoch": 32.31597323247018, + "grad_norm": 0.12983249127864838, + "learning_rate": 4.520204787702181e-07, + "loss": 0.0095, + "step": 111070 + }, + { + "epoch": 32.31888274658132, + "grad_norm": 0.21211929619312286, + "learning_rate": 4.5101389553899476e-07, + "loss": 0.0115, + "step": 111080 + }, + { + "epoch": 32.32179226069246, + "grad_norm": 0.11572299897670746, + "learning_rate": 4.500084172072499e-07, + "loss": 0.01, + "step": 111090 + }, + { + "epoch": 32.32470177480361, + "grad_norm": 0.07481752336025238, + "learning_rate": 4.4900404385134266e-07, + "loss": 0.009, + "step": 111100 + }, + { + "epoch": 32.32761128891475, + "grad_norm": 0.16645170748233795, + "learning_rate": 4.4800077554755015e-07, + "loss": 0.0097, + "step": 111110 + }, + { + "epoch": 32.3305208030259, + "grad_norm": 0.11020021885633469, + "learning_rate": 4.4699861237206654e-07, + "loss": 0.01, + "step": 111120 + }, + { + "epoch": 32.33343031713704, + "grad_norm": 0.17686229944229126, + "learning_rate": 4.459975544009942e-07, + "loss": 0.0126, + "step": 111130 + }, + { + "epoch": 32.33633983124818, + "grad_norm": 0.09133900701999664, + "learning_rate": 4.4499760171036075e-07, + "loss": 0.0103, + "step": 111140 + }, + { + "epoch": 32.33924934535933, + "grad_norm": 0.12554562091827393, + "learning_rate": 4.4399875437610374e-07, + "loss": 0.0108, + "step": 111150 + }, + { + "epoch": 32.342158859470466, + "grad_norm": 0.12662529945373535, + "learning_rate": 4.4300101247408074e-07, + "loss": 0.0091, + "step": 111160 + }, + { + "epoch": 32.34506837358161, + "grad_norm": 0.07981674373149872, + "learning_rate": 4.4200437608006294e-07, + "loss": 0.0102, + "step": 111170 + }, + { + "epoch": 32.34797788769276, + "grad_norm": 0.15082454681396484, + "learning_rate": 4.4100884526973806e-07, + "loss": 0.0101, + "step": 111180 + }, + { + "epoch": 32.3508874018039, + "grad_norm": 0.11536887288093567, + "learning_rate": 4.400144201187123e-07, + "loss": 0.011, + "step": 111190 + }, + { + "epoch": 32.35379691591504, + "grad_norm": 0.11873277276754379, + "learning_rate": 4.390211007025002e-07, + "loss": 0.0096, + "step": 111200 + }, + { + "epoch": 32.356706430026186, + "grad_norm": 0.12898659706115723, + "learning_rate": 4.3802888709654487e-07, + "loss": 0.0082, + "step": 111210 + }, + { + "epoch": 32.35961594413733, + "grad_norm": 0.09966445714235306, + "learning_rate": 4.3703777937619263e-07, + "loss": 0.0105, + "step": 111220 + }, + { + "epoch": 32.36252545824847, + "grad_norm": 0.12183159589767456, + "learning_rate": 4.3604777761671487e-07, + "loss": 0.0103, + "step": 111230 + }, + { + "epoch": 32.365434972359616, + "grad_norm": 0.08176948130130768, + "learning_rate": 4.350588818932949e-07, + "loss": 0.0093, + "step": 111240 + }, + { + "epoch": 32.36834448647076, + "grad_norm": 0.09401549398899078, + "learning_rate": 4.340710922810326e-07, + "loss": 0.0094, + "step": 111250 + }, + { + "epoch": 32.3712540005819, + "grad_norm": 0.09176450222730637, + "learning_rate": 4.330844088549413e-07, + "loss": 0.0111, + "step": 111260 + }, + { + "epoch": 32.374163514693045, + "grad_norm": 0.12290462851524353, + "learning_rate": 4.320988316899593e-07, + "loss": 0.0088, + "step": 111270 + }, + { + "epoch": 32.37707302880419, + "grad_norm": 0.1038379818201065, + "learning_rate": 4.311143608609319e-07, + "loss": 0.0112, + "step": 111280 + }, + { + "epoch": 32.379982542915336, + "grad_norm": 0.11723941564559937, + "learning_rate": 4.301309964426192e-07, + "loss": 0.0098, + "step": 111290 + }, + { + "epoch": 32.382892057026474, + "grad_norm": 0.15048016607761383, + "learning_rate": 4.291487385097065e-07, + "loss": 0.0108, + "step": 111300 + }, + { + "epoch": 32.38580157113762, + "grad_norm": 0.13045470416545868, + "learning_rate": 4.2816758713678903e-07, + "loss": 0.0127, + "step": 111310 + }, + { + "epoch": 32.388711085248765, + "grad_norm": 0.12080246210098267, + "learning_rate": 4.2718754239837567e-07, + "loss": 0.0105, + "step": 111320 + }, + { + "epoch": 32.391620599359904, + "grad_norm": 0.06512724608182907, + "learning_rate": 4.2620860436889674e-07, + "loss": 0.0096, + "step": 111330 + }, + { + "epoch": 32.39453011347105, + "grad_norm": 0.08803107589483261, + "learning_rate": 4.252307731226945e-07, + "loss": 0.0101, + "step": 111340 + }, + { + "epoch": 32.397439627582195, + "grad_norm": 0.11606394499540329, + "learning_rate": 4.2425404873403295e-07, + "loss": 0.0102, + "step": 111350 + }, + { + "epoch": 32.40034914169334, + "grad_norm": 0.08336324989795685, + "learning_rate": 4.232784312770827e-07, + "loss": 0.0093, + "step": 111360 + }, + { + "epoch": 32.40325865580448, + "grad_norm": 0.06973573565483093, + "learning_rate": 4.2230392082593605e-07, + "loss": 0.0091, + "step": 111370 + }, + { + "epoch": 32.406168169915624, + "grad_norm": 0.08586416393518448, + "learning_rate": 4.213305174546056e-07, + "loss": 0.0088, + "step": 111380 + }, + { + "epoch": 32.40907768402677, + "grad_norm": 0.11824632436037064, + "learning_rate": 4.203582212370105e-07, + "loss": 0.0117, + "step": 111390 + }, + { + "epoch": 32.41198719813791, + "grad_norm": 0.08724639564752579, + "learning_rate": 4.1938703224699e-07, + "loss": 0.0103, + "step": 111400 + }, + { + "epoch": 32.41489671224905, + "grad_norm": 0.10601670295000076, + "learning_rate": 4.184169505583002e-07, + "loss": 0.0143, + "step": 111410 + }, + { + "epoch": 32.4178062263602, + "grad_norm": 0.09941036254167557, + "learning_rate": 4.1744797624461537e-07, + "loss": 0.0111, + "step": 111420 + }, + { + "epoch": 32.420715740471344, + "grad_norm": 0.09278449416160583, + "learning_rate": 4.164801093795184e-07, + "loss": 0.0099, + "step": 111430 + }, + { + "epoch": 32.42362525458248, + "grad_norm": 0.09807746857404709, + "learning_rate": 4.1551335003651545e-07, + "loss": 0.0119, + "step": 111440 + }, + { + "epoch": 32.42653476869363, + "grad_norm": 0.08589839935302734, + "learning_rate": 4.145476982890228e-07, + "loss": 0.01, + "step": 111450 + }, + { + "epoch": 32.42944428280477, + "grad_norm": 0.12116020917892456, + "learning_rate": 4.135831542103785e-07, + "loss": 0.0107, + "step": 111460 + }, + { + "epoch": 32.43235379691591, + "grad_norm": 0.1271299570798874, + "learning_rate": 4.1261971787383056e-07, + "loss": 0.0105, + "step": 111470 + }, + { + "epoch": 32.43526331102706, + "grad_norm": 0.10649609565734863, + "learning_rate": 4.1165738935254716e-07, + "loss": 0.0098, + "step": 111480 + }, + { + "epoch": 32.4381728251382, + "grad_norm": 0.10874376446008682, + "learning_rate": 4.1069616871961147e-07, + "loss": 0.0117, + "step": 111490 + }, + { + "epoch": 32.44108233924935, + "grad_norm": 0.12766917049884796, + "learning_rate": 4.0973605604801846e-07, + "loss": 0.0103, + "step": 111500 + }, + { + "epoch": 32.44399185336049, + "grad_norm": 0.12993252277374268, + "learning_rate": 4.0877705141068645e-07, + "loss": 0.0099, + "step": 111510 + }, + { + "epoch": 32.44690136747163, + "grad_norm": 0.1120728850364685, + "learning_rate": 4.078191548804422e-07, + "loss": 0.0107, + "step": 111520 + }, + { + "epoch": 32.44981088158278, + "grad_norm": 0.10824897140264511, + "learning_rate": 4.068623665300342e-07, + "loss": 0.0117, + "step": 111530 + }, + { + "epoch": 32.452720395693916, + "grad_norm": 0.10666760802268982, + "learning_rate": 4.059066864321226e-07, + "loss": 0.01, + "step": 111540 + }, + { + "epoch": 32.45562990980506, + "grad_norm": 0.10978882759809494, + "learning_rate": 4.049521146592877e-07, + "loss": 0.012, + "step": 111550 + }, + { + "epoch": 32.45853942391621, + "grad_norm": 0.12022983282804489, + "learning_rate": 4.0399865128401824e-07, + "loss": 0.0113, + "step": 111560 + }, + { + "epoch": 32.46144893802735, + "grad_norm": 0.12384084612131119, + "learning_rate": 4.030462963787246e-07, + "loss": 0.0097, + "step": 111570 + }, + { + "epoch": 32.46435845213849, + "grad_norm": 0.1210019513964653, + "learning_rate": 4.020950500157372e-07, + "loss": 0.0093, + "step": 111580 + }, + { + "epoch": 32.467267966249636, + "grad_norm": 0.0940258800983429, + "learning_rate": 4.0114491226729e-07, + "loss": 0.0108, + "step": 111590 + }, + { + "epoch": 32.47017748036078, + "grad_norm": 0.11296091973781586, + "learning_rate": 4.0019588320554355e-07, + "loss": 0.0109, + "step": 111600 + }, + { + "epoch": 32.47308699447192, + "grad_norm": 0.09578882157802582, + "learning_rate": 3.9924796290257025e-07, + "loss": 0.0098, + "step": 111610 + }, + { + "epoch": 32.475996508583066, + "grad_norm": 0.084278903901577, + "learning_rate": 3.983011514303575e-07, + "loss": 0.0089, + "step": 111620 + }, + { + "epoch": 32.47890602269421, + "grad_norm": 0.10463351756334305, + "learning_rate": 3.973554488608078e-07, + "loss": 0.0092, + "step": 111630 + }, + { + "epoch": 32.48181553680536, + "grad_norm": 0.08153869956731796, + "learning_rate": 3.9641085526574364e-07, + "loss": 0.0108, + "step": 111640 + }, + { + "epoch": 32.484725050916495, + "grad_norm": 0.1398952603340149, + "learning_rate": 3.9546737071690097e-07, + "loss": 0.0106, + "step": 111650 + }, + { + "epoch": 32.48763456502764, + "grad_norm": 0.10244506597518921, + "learning_rate": 3.9452499528592924e-07, + "loss": 0.0086, + "step": 111660 + }, + { + "epoch": 32.490544079138786, + "grad_norm": 0.12245788425207138, + "learning_rate": 3.9358372904439277e-07, + "loss": 0.0102, + "step": 111670 + }, + { + "epoch": 32.493453593249924, + "grad_norm": 0.214147686958313, + "learning_rate": 3.926435720637828e-07, + "loss": 0.0113, + "step": 111680 + }, + { + "epoch": 32.49636310736107, + "grad_norm": 0.1188432052731514, + "learning_rate": 3.917045244154921e-07, + "loss": 0.01, + "step": 111690 + }, + { + "epoch": 32.499272621472215, + "grad_norm": 0.14390525221824646, + "learning_rate": 3.907665861708354e-07, + "loss": 0.0087, + "step": 111700 + }, + { + "epoch": 32.50218213558336, + "grad_norm": 0.130576491355896, + "learning_rate": 3.89829757401044e-07, + "loss": 0.01, + "step": 111710 + }, + { + "epoch": 32.5050916496945, + "grad_norm": 0.1217254102230072, + "learning_rate": 3.8889403817726596e-07, + "loss": 0.0121, + "step": 111720 + }, + { + "epoch": 32.508001163805645, + "grad_norm": 0.08308164030313492, + "learning_rate": 3.8795942857055957e-07, + "loss": 0.0098, + "step": 111730 + }, + { + "epoch": 32.51091067791679, + "grad_norm": 0.09402818977832794, + "learning_rate": 3.870259286519029e-07, + "loss": 0.0102, + "step": 111740 + }, + { + "epoch": 32.51382019202793, + "grad_norm": 0.13535046577453613, + "learning_rate": 3.860935384921893e-07, + "loss": 0.0093, + "step": 111750 + }, + { + "epoch": 32.516729706139074, + "grad_norm": 0.09611926227807999, + "learning_rate": 3.851622581622305e-07, + "loss": 0.0113, + "step": 111760 + }, + { + "epoch": 32.51963922025022, + "grad_norm": 0.15040569007396698, + "learning_rate": 3.8423208773274654e-07, + "loss": 0.0111, + "step": 111770 + }, + { + "epoch": 32.522548734361365, + "grad_norm": 0.09004873037338257, + "learning_rate": 3.8330302727437924e-07, + "loss": 0.0114, + "step": 111780 + }, + { + "epoch": 32.5254582484725, + "grad_norm": 0.21302726864814758, + "learning_rate": 3.8237507685768724e-07, + "loss": 0.0089, + "step": 111790 + }, + { + "epoch": 32.52836776258365, + "grad_norm": 0.12573082745075226, + "learning_rate": 3.814482365531391e-07, + "loss": 0.0092, + "step": 111800 + }, + { + "epoch": 32.531277276694794, + "grad_norm": 0.11062635481357574, + "learning_rate": 3.805225064311235e-07, + "loss": 0.0107, + "step": 111810 + }, + { + "epoch": 32.53418679080593, + "grad_norm": 0.08884147554636002, + "learning_rate": 3.795978865619426e-07, + "loss": 0.0094, + "step": 111820 + }, + { + "epoch": 32.53709630491708, + "grad_norm": 0.11463578790426254, + "learning_rate": 3.7867437701581854e-07, + "loss": 0.01, + "step": 111830 + }, + { + "epoch": 32.540005819028224, + "grad_norm": 0.19807176291942596, + "learning_rate": 3.7775197786288017e-07, + "loss": 0.0095, + "step": 111840 + }, + { + "epoch": 32.54291533313936, + "grad_norm": 0.12507747113704681, + "learning_rate": 3.7683068917318317e-07, + "loss": 0.0088, + "step": 111850 + }, + { + "epoch": 32.54582484725051, + "grad_norm": 0.11582974344491959, + "learning_rate": 3.759105110166883e-07, + "loss": 0.0105, + "step": 111860 + }, + { + "epoch": 32.54873436136165, + "grad_norm": 0.05433482676744461, + "learning_rate": 3.7499144346328117e-07, + "loss": 0.0088, + "step": 111870 + }, + { + "epoch": 32.5516438754728, + "grad_norm": 0.1670781970024109, + "learning_rate": 3.740734865827561e-07, + "loss": 0.0093, + "step": 111880 + }, + { + "epoch": 32.55455338958394, + "grad_norm": 0.11096125096082687, + "learning_rate": 3.731566404448272e-07, + "loss": 0.01, + "step": 111890 + }, + { + "epoch": 32.55746290369508, + "grad_norm": 0.13282544910907745, + "learning_rate": 3.722409051191239e-07, + "loss": 0.0096, + "step": 111900 + }, + { + "epoch": 32.56037241780623, + "grad_norm": 0.09100012481212616, + "learning_rate": 3.713262806751888e-07, + "loss": 0.0102, + "step": 111910 + }, + { + "epoch": 32.563281931917366, + "grad_norm": 0.12441157549619675, + "learning_rate": 3.704127671824814e-07, + "loss": 0.0112, + "step": 111920 + }, + { + "epoch": 32.56619144602851, + "grad_norm": 0.13893058896064758, + "learning_rate": 3.6950036471037786e-07, + "loss": 0.0118, + "step": 111930 + }, + { + "epoch": 32.56910096013966, + "grad_norm": 0.0781932920217514, + "learning_rate": 3.685890733281677e-07, + "loss": 0.0104, + "step": 111940 + }, + { + "epoch": 32.5720104742508, + "grad_norm": 0.13036972284317017, + "learning_rate": 3.6767889310505895e-07, + "loss": 0.0104, + "step": 111950 + }, + { + "epoch": 32.57491998836194, + "grad_norm": 0.10559894889593124, + "learning_rate": 3.667698241101747e-07, + "loss": 0.0095, + "step": 111960 + }, + { + "epoch": 32.57782950247309, + "grad_norm": 0.10507365316152573, + "learning_rate": 3.6586186641254794e-07, + "loss": 0.0113, + "step": 111970 + }, + { + "epoch": 32.58073901658423, + "grad_norm": 0.10218793153762817, + "learning_rate": 3.6495502008113855e-07, + "loss": 0.0074, + "step": 111980 + }, + { + "epoch": 32.58364853069537, + "grad_norm": 0.07106108218431473, + "learning_rate": 3.640492851848132e-07, + "loss": 0.0123, + "step": 111990 + }, + { + "epoch": 32.586558044806516, + "grad_norm": 0.0677674412727356, + "learning_rate": 3.631446617923534e-07, + "loss": 0.0097, + "step": 112000 + }, + { + "epoch": 32.58946755891766, + "grad_norm": 0.0977654904127121, + "learning_rate": 3.622411499724626e-07, + "loss": 0.0102, + "step": 112010 + }, + { + "epoch": 32.59237707302881, + "grad_norm": 0.09729558229446411, + "learning_rate": 3.613387497937559e-07, + "loss": 0.0115, + "step": 112020 + }, + { + "epoch": 32.595286587139945, + "grad_norm": 0.12168290466070175, + "learning_rate": 3.604374613247652e-07, + "loss": 0.0101, + "step": 112030 + }, + { + "epoch": 32.59819610125109, + "grad_norm": 0.07819133996963501, + "learning_rate": 3.5953728463393565e-07, + "loss": 0.0118, + "step": 112040 + }, + { + "epoch": 32.601105615362236, + "grad_norm": 0.17314469814300537, + "learning_rate": 3.586382197896293e-07, + "loss": 0.0136, + "step": 112050 + }, + { + "epoch": 32.604015129473375, + "grad_norm": 0.14231617748737335, + "learning_rate": 3.5774026686012817e-07, + "loss": 0.009, + "step": 112060 + }, + { + "epoch": 32.60692464358452, + "grad_norm": 0.1770656704902649, + "learning_rate": 3.5684342591362106e-07, + "loss": 0.0113, + "step": 112070 + }, + { + "epoch": 32.609834157695666, + "grad_norm": 0.08865339308977127, + "learning_rate": 3.559476970182185e-07, + "loss": 0.0101, + "step": 112080 + }, + { + "epoch": 32.61274367180681, + "grad_norm": 0.10371583700180054, + "learning_rate": 3.550530802419494e-07, + "loss": 0.0113, + "step": 112090 + }, + { + "epoch": 32.61565318591795, + "grad_norm": 0.07978455722332001, + "learning_rate": 3.541595756527477e-07, + "loss": 0.0098, + "step": 112100 + }, + { + "epoch": 32.618562700029095, + "grad_norm": 0.11771667003631592, + "learning_rate": 3.532671833184742e-07, + "loss": 0.0105, + "step": 112110 + }, + { + "epoch": 32.62147221414024, + "grad_norm": 0.1567104160785675, + "learning_rate": 3.5237590330689793e-07, + "loss": 0.0111, + "step": 112120 + }, + { + "epoch": 32.62438172825138, + "grad_norm": 0.08990521728992462, + "learning_rate": 3.5148573568570485e-07, + "loss": 0.0121, + "step": 112130 + }, + { + "epoch": 32.627291242362524, + "grad_norm": 0.09072469174861908, + "learning_rate": 3.5059668052249916e-07, + "loss": 0.0093, + "step": 112140 + }, + { + "epoch": 32.63020075647367, + "grad_norm": 0.09988945722579956, + "learning_rate": 3.497087378847985e-07, + "loss": 0.0099, + "step": 112150 + }, + { + "epoch": 32.633110270584815, + "grad_norm": 0.10056976228952408, + "learning_rate": 3.4882190784003397e-07, + "loss": 0.0099, + "step": 112160 + }, + { + "epoch": 32.636019784695954, + "grad_norm": 0.08479641377925873, + "learning_rate": 3.4793619045555505e-07, + "loss": 0.0081, + "step": 112170 + }, + { + "epoch": 32.6389292988071, + "grad_norm": 0.10954535752534866, + "learning_rate": 3.4705158579863116e-07, + "loss": 0.0102, + "step": 112180 + }, + { + "epoch": 32.641838812918245, + "grad_norm": 0.15672831237316132, + "learning_rate": 3.461680939364353e-07, + "loss": 0.0109, + "step": 112190 + }, + { + "epoch": 32.64474832702938, + "grad_norm": 0.08521413803100586, + "learning_rate": 3.452857149360672e-07, + "loss": 0.009, + "step": 112200 + }, + { + "epoch": 32.64765784114053, + "grad_norm": 0.11338034272193909, + "learning_rate": 3.444044488645348e-07, + "loss": 0.0095, + "step": 112210 + }, + { + "epoch": 32.650567355251674, + "grad_norm": 0.12237349152565002, + "learning_rate": 3.43524295788768e-07, + "loss": 0.0097, + "step": 112220 + }, + { + "epoch": 32.65347686936282, + "grad_norm": 0.12255679816007614, + "learning_rate": 3.426452557756066e-07, + "loss": 0.0097, + "step": 112230 + }, + { + "epoch": 32.65638638347396, + "grad_norm": 0.0834449976682663, + "learning_rate": 3.417673288918055e-07, + "loss": 0.0101, + "step": 112240 + }, + { + "epoch": 32.6592958975851, + "grad_norm": 0.10250643640756607, + "learning_rate": 3.4089051520404144e-07, + "loss": 0.0108, + "step": 112250 + }, + { + "epoch": 32.66220541169625, + "grad_norm": 0.137957364320755, + "learning_rate": 3.4001481477890284e-07, + "loss": 0.0107, + "step": 112260 + }, + { + "epoch": 32.66511492580739, + "grad_norm": 0.09586115181446075, + "learning_rate": 3.391402276828881e-07, + "loss": 0.0086, + "step": 112270 + }, + { + "epoch": 32.66802443991853, + "grad_norm": 0.07544173300266266, + "learning_rate": 3.382667539824225e-07, + "loss": 0.0104, + "step": 112280 + }, + { + "epoch": 32.67093395402968, + "grad_norm": 0.12036170810461044, + "learning_rate": 3.373943937438362e-07, + "loss": 0.0091, + "step": 112290 + }, + { + "epoch": 32.673843468140824, + "grad_norm": 0.08929306268692017, + "learning_rate": 3.365231470333813e-07, + "loss": 0.0106, + "step": 112300 + }, + { + "epoch": 32.67675298225196, + "grad_norm": 0.16525638103485107, + "learning_rate": 3.356530139172231e-07, + "loss": 0.0094, + "step": 112310 + }, + { + "epoch": 32.67966249636311, + "grad_norm": 0.13282091915607452, + "learning_rate": 3.3478399446144223e-07, + "loss": 0.0094, + "step": 112320 + }, + { + "epoch": 32.68257201047425, + "grad_norm": 0.058094579726457596, + "learning_rate": 3.3391608873203585e-07, + "loss": 0.0084, + "step": 112330 + }, + { + "epoch": 32.68548152458539, + "grad_norm": 0.09103244543075562, + "learning_rate": 3.3304929679491445e-07, + "loss": 0.0094, + "step": 112340 + }, + { + "epoch": 32.68839103869654, + "grad_norm": 0.07913636416196823, + "learning_rate": 3.321836187159039e-07, + "loss": 0.011, + "step": 112350 + }, + { + "epoch": 32.69130055280768, + "grad_norm": 0.09530390799045563, + "learning_rate": 3.313190545607497e-07, + "loss": 0.0096, + "step": 112360 + }, + { + "epoch": 32.69421006691883, + "grad_norm": 0.12582288682460785, + "learning_rate": 3.3045560439510794e-07, + "loss": 0.0087, + "step": 112370 + }, + { + "epoch": 32.697119581029966, + "grad_norm": 0.0548134371638298, + "learning_rate": 3.295932682845526e-07, + "loss": 0.0101, + "step": 112380 + }, + { + "epoch": 32.70002909514111, + "grad_norm": 0.09979178011417389, + "learning_rate": 3.287320462945731e-07, + "loss": 0.0095, + "step": 112390 + }, + { + "epoch": 32.70293860925226, + "grad_norm": 0.10425344109535217, + "learning_rate": 3.278719384905721e-07, + "loss": 0.0121, + "step": 112400 + }, + { + "epoch": 32.705848123363396, + "grad_norm": 0.11888730525970459, + "learning_rate": 3.270129449378689e-07, + "loss": 0.0107, + "step": 112410 + }, + { + "epoch": 32.70875763747454, + "grad_norm": 0.08825445175170898, + "learning_rate": 3.261550657016998e-07, + "loss": 0.0099, + "step": 112420 + }, + { + "epoch": 32.71166715158569, + "grad_norm": 0.08831053227186203, + "learning_rate": 3.25298300847216e-07, + "loss": 0.0096, + "step": 112430 + }, + { + "epoch": 32.71457666569683, + "grad_norm": 0.08775255084037781, + "learning_rate": 3.2444265043947875e-07, + "loss": 0.0096, + "step": 112440 + }, + { + "epoch": 32.71748617980797, + "grad_norm": 0.11735421419143677, + "learning_rate": 3.235881145434744e-07, + "loss": 0.01, + "step": 112450 + }, + { + "epoch": 32.720395693919116, + "grad_norm": 0.11075551062822342, + "learning_rate": 3.2273469322409597e-07, + "loss": 0.0104, + "step": 112460 + }, + { + "epoch": 32.72330520803026, + "grad_norm": 0.11101174354553223, + "learning_rate": 3.218823865461551e-07, + "loss": 0.0102, + "step": 112470 + }, + { + "epoch": 32.7262147221414, + "grad_norm": 0.13197682797908783, + "learning_rate": 3.2103119457438156e-07, + "loss": 0.0102, + "step": 112480 + }, + { + "epoch": 32.729124236252545, + "grad_norm": 0.10370247811079025, + "learning_rate": 3.201811173734154e-07, + "loss": 0.0096, + "step": 112490 + }, + { + "epoch": 32.73203375036369, + "grad_norm": 0.11820723116397858, + "learning_rate": 3.1933215500781486e-07, + "loss": 0.0111, + "step": 112500 + }, + { + "epoch": 32.734943264474836, + "grad_norm": 0.1326059252023697, + "learning_rate": 3.184843075420535e-07, + "loss": 0.0082, + "step": 112510 + }, + { + "epoch": 32.737852778585975, + "grad_norm": 0.09973543137311935, + "learning_rate": 3.176375750405197e-07, + "loss": 0.009, + "step": 112520 + }, + { + "epoch": 32.74076229269712, + "grad_norm": 0.09740941971540451, + "learning_rate": 3.1679195756751543e-07, + "loss": 0.0097, + "step": 112530 + }, + { + "epoch": 32.743671806808266, + "grad_norm": 0.20631635189056396, + "learning_rate": 3.15947455187261e-07, + "loss": 0.0122, + "step": 112540 + }, + { + "epoch": 32.746581320919404, + "grad_norm": 0.09007777273654938, + "learning_rate": 3.1510406796389336e-07, + "loss": 0.0088, + "step": 112550 + }, + { + "epoch": 32.74949083503055, + "grad_norm": 0.12285393476486206, + "learning_rate": 3.1426179596145964e-07, + "loss": 0.0134, + "step": 112560 + }, + { + "epoch": 32.752400349141695, + "grad_norm": 0.12728196382522583, + "learning_rate": 3.1342063924392373e-07, + "loss": 0.0116, + "step": 112570 + }, + { + "epoch": 32.75530986325283, + "grad_norm": 0.13467590510845184, + "learning_rate": 3.125805978751678e-07, + "loss": 0.0088, + "step": 112580 + }, + { + "epoch": 32.75821937736398, + "grad_norm": 0.09037640690803528, + "learning_rate": 3.117416719189875e-07, + "loss": 0.0085, + "step": 112590 + }, + { + "epoch": 32.761128891475124, + "grad_norm": 0.3208988904953003, + "learning_rate": 3.1090386143909186e-07, + "loss": 0.009, + "step": 112600 + }, + { + "epoch": 32.76403840558627, + "grad_norm": 0.10400837659835815, + "learning_rate": 3.1006716649910835e-07, + "loss": 0.0099, + "step": 112610 + }, + { + "epoch": 32.76694791969741, + "grad_norm": 0.11001506447792053, + "learning_rate": 3.092315871625795e-07, + "loss": 0.0108, + "step": 112620 + }, + { + "epoch": 32.76985743380855, + "grad_norm": 0.08171840757131577, + "learning_rate": 3.0839712349296113e-07, + "loss": 0.0077, + "step": 112630 + }, + { + "epoch": 32.7727669479197, + "grad_norm": 0.07080948352813721, + "learning_rate": 3.0756377555362425e-07, + "loss": 0.007, + "step": 112640 + }, + { + "epoch": 32.77567646203084, + "grad_norm": 0.08289693295955658, + "learning_rate": 3.0673154340785656e-07, + "loss": 0.0102, + "step": 112650 + }, + { + "epoch": 32.77858597614198, + "grad_norm": 0.12500505149364471, + "learning_rate": 3.0590042711886257e-07, + "loss": 0.009, + "step": 112660 + }, + { + "epoch": 32.78149549025313, + "grad_norm": 0.09072291105985641, + "learning_rate": 3.050704267497584e-07, + "loss": 0.0105, + "step": 112670 + }, + { + "epoch": 32.784405004364274, + "grad_norm": 0.11598008871078491, + "learning_rate": 3.042415423635736e-07, + "loss": 0.0093, + "step": 112680 + }, + { + "epoch": 32.78731451847541, + "grad_norm": 0.0889853984117508, + "learning_rate": 3.034137740232645e-07, + "loss": 0.0105, + "step": 112690 + }, + { + "epoch": 32.79022403258656, + "grad_norm": 0.09205038100481033, + "learning_rate": 3.0258712179168924e-07, + "loss": 0.0081, + "step": 112700 + }, + { + "epoch": 32.7931335466977, + "grad_norm": 0.15595465898513794, + "learning_rate": 3.017615857316258e-07, + "loss": 0.0102, + "step": 112710 + }, + { + "epoch": 32.79604306080884, + "grad_norm": 0.10932721942663193, + "learning_rate": 3.009371659057708e-07, + "loss": 0.0111, + "step": 112720 + }, + { + "epoch": 32.79895257491999, + "grad_norm": 0.14322657883167267, + "learning_rate": 3.001138623767341e-07, + "loss": 0.0113, + "step": 112730 + }, + { + "epoch": 32.80186208903113, + "grad_norm": 0.11328393965959549, + "learning_rate": 2.9929167520703737e-07, + "loss": 0.0103, + "step": 112740 + }, + { + "epoch": 32.80477160314228, + "grad_norm": 0.09843549877405167, + "learning_rate": 2.984706044591223e-07, + "loss": 0.0121, + "step": 112750 + }, + { + "epoch": 32.807681117253416, + "grad_norm": 0.08228198438882828, + "learning_rate": 2.976506501953424e-07, + "loss": 0.0097, + "step": 112760 + }, + { + "epoch": 32.81059063136456, + "grad_norm": 0.12125616520643234, + "learning_rate": 2.968318124779695e-07, + "loss": 0.0094, + "step": 112770 + }, + { + "epoch": 32.81350014547571, + "grad_norm": 0.1638822704553604, + "learning_rate": 2.9601409136918723e-07, + "loss": 0.0091, + "step": 112780 + }, + { + "epoch": 32.816409659586846, + "grad_norm": 0.1122971773147583, + "learning_rate": 2.951974869310958e-07, + "loss": 0.0094, + "step": 112790 + }, + { + "epoch": 32.81931917369799, + "grad_norm": 0.08341007679700851, + "learning_rate": 2.943819992257141e-07, + "loss": 0.0099, + "step": 112800 + }, + { + "epoch": 32.82222868780914, + "grad_norm": 0.0893196314573288, + "learning_rate": 2.9356762831497076e-07, + "loss": 0.0092, + "step": 112810 + }, + { + "epoch": 32.82513820192028, + "grad_norm": 0.09626497328281403, + "learning_rate": 2.927543742607114e-07, + "loss": 0.0102, + "step": 112820 + }, + { + "epoch": 32.82804771603142, + "grad_norm": 0.09905099123716354, + "learning_rate": 2.9194223712469825e-07, + "loss": 0.0098, + "step": 112830 + }, + { + "epoch": 32.830957230142566, + "grad_norm": 0.10025909543037415, + "learning_rate": 2.9113121696860537e-07, + "loss": 0.0112, + "step": 112840 + }, + { + "epoch": 32.83386674425371, + "grad_norm": 0.105056531727314, + "learning_rate": 2.9032131385403004e-07, + "loss": 0.0107, + "step": 112850 + }, + { + "epoch": 32.83677625836485, + "grad_norm": 0.08639685064554214, + "learning_rate": 2.8951252784247317e-07, + "loss": 0.0102, + "step": 112860 + }, + { + "epoch": 32.839685772475995, + "grad_norm": 0.08905529230833054, + "learning_rate": 2.8870485899535893e-07, + "loss": 0.0128, + "step": 112870 + }, + { + "epoch": 32.84259528658714, + "grad_norm": 0.11838354170322418, + "learning_rate": 2.878983073740249e-07, + "loss": 0.0095, + "step": 112880 + }, + { + "epoch": 32.845504800698286, + "grad_norm": 0.11335836350917816, + "learning_rate": 2.8709287303972373e-07, + "loss": 0.0087, + "step": 112890 + }, + { + "epoch": 32.848414314809425, + "grad_norm": 0.12598459422588348, + "learning_rate": 2.862885560536216e-07, + "loss": 0.0092, + "step": 112900 + }, + { + "epoch": 32.85132382892057, + "grad_norm": 0.1186632290482521, + "learning_rate": 2.854853564768012e-07, + "loss": 0.0109, + "step": 112910 + }, + { + "epoch": 32.854233343031716, + "grad_norm": 0.08748916536569595, + "learning_rate": 2.846832743702621e-07, + "loss": 0.0091, + "step": 112920 + }, + { + "epoch": 32.857142857142854, + "grad_norm": 0.08060594648122787, + "learning_rate": 2.838823097949156e-07, + "loss": 0.011, + "step": 112930 + }, + { + "epoch": 32.860052371254, + "grad_norm": 0.14499728381633759, + "learning_rate": 2.830824628115863e-07, + "loss": 0.0108, + "step": 112940 + }, + { + "epoch": 32.862961885365145, + "grad_norm": 0.11563021689653397, + "learning_rate": 2.822837334810241e-07, + "loss": 0.0103, + "step": 112950 + }, + { + "epoch": 32.86587139947629, + "grad_norm": 0.08188991993665695, + "learning_rate": 2.814861218638837e-07, + "loss": 0.0105, + "step": 112960 + }, + { + "epoch": 32.86878091358743, + "grad_norm": 0.12633419036865234, + "learning_rate": 2.8068962802073995e-07, + "loss": 0.0083, + "step": 112970 + }, + { + "epoch": 32.871690427698574, + "grad_norm": 0.09361465275287628, + "learning_rate": 2.7989425201207453e-07, + "loss": 0.0122, + "step": 112980 + }, + { + "epoch": 32.87459994180972, + "grad_norm": 0.195834219455719, + "learning_rate": 2.7909999389830076e-07, + "loss": 0.0098, + "step": 112990 + }, + { + "epoch": 32.87750945592086, + "grad_norm": 0.10482145845890045, + "learning_rate": 2.7830685373973375e-07, + "loss": 0.0103, + "step": 113000 + }, + { + "epoch": 32.880418970032004, + "grad_norm": 0.10053907334804535, + "learning_rate": 2.7751483159660364e-07, + "loss": 0.0099, + "step": 113010 + }, + { + "epoch": 32.88332848414315, + "grad_norm": 0.12031325697898865, + "learning_rate": 2.767239275290623e-07, + "loss": 0.0101, + "step": 113020 + }, + { + "epoch": 32.886237998254295, + "grad_norm": 0.18275347352027893, + "learning_rate": 2.75934141597175e-07, + "loss": 0.0104, + "step": 113030 + }, + { + "epoch": 32.88914751236543, + "grad_norm": 0.0899670273065567, + "learning_rate": 2.751454738609205e-07, + "loss": 0.0101, + "step": 113040 + }, + { + "epoch": 32.89205702647658, + "grad_norm": 0.11232806742191315, + "learning_rate": 2.7435792438018747e-07, + "loss": 0.0093, + "step": 113050 + }, + { + "epoch": 32.894966540587724, + "grad_norm": 0.11384978145360947, + "learning_rate": 2.7357149321479315e-07, + "loss": 0.0108, + "step": 113060 + }, + { + "epoch": 32.89787605469886, + "grad_norm": 0.09309344738721848, + "learning_rate": 2.7278618042445637e-07, + "loss": 0.0111, + "step": 113070 + }, + { + "epoch": 32.90078556881001, + "grad_norm": 0.09238001704216003, + "learning_rate": 2.720019860688178e-07, + "loss": 0.0081, + "step": 113080 + }, + { + "epoch": 32.90369508292115, + "grad_norm": 0.21325984597206116, + "learning_rate": 2.7121891020743305e-07, + "loss": 0.0096, + "step": 113090 + }, + { + "epoch": 32.9066045970323, + "grad_norm": 0.10024705529212952, + "learning_rate": 2.704369528997697e-07, + "loss": 0.0105, + "step": 113100 + }, + { + "epoch": 32.90951411114344, + "grad_norm": 0.07796116918325424, + "learning_rate": 2.696561142052134e-07, + "loss": 0.0095, + "step": 113110 + }, + { + "epoch": 32.91242362525458, + "grad_norm": 0.15487992763519287, + "learning_rate": 2.688763941830635e-07, + "loss": 0.0097, + "step": 113120 + }, + { + "epoch": 32.91533313936573, + "grad_norm": 0.10542180389165878, + "learning_rate": 2.6809779289253435e-07, + "loss": 0.011, + "step": 113130 + }, + { + "epoch": 32.91824265347687, + "grad_norm": 0.12981706857681274, + "learning_rate": 2.6732031039275684e-07, + "loss": 0.0127, + "step": 113140 + }, + { + "epoch": 32.92115216758801, + "grad_norm": 0.12542155385017395, + "learning_rate": 2.6654394674277216e-07, + "loss": 0.0126, + "step": 113150 + }, + { + "epoch": 32.92406168169916, + "grad_norm": 0.13123798370361328, + "learning_rate": 2.657687020015431e-07, + "loss": 0.0115, + "step": 113160 + }, + { + "epoch": 32.9269711958103, + "grad_norm": 0.15597540140151978, + "learning_rate": 2.6499457622794596e-07, + "loss": 0.0088, + "step": 113170 + }, + { + "epoch": 32.92988070992144, + "grad_norm": 0.14452043175697327, + "learning_rate": 2.6422156948076536e-07, + "loss": 0.0107, + "step": 113180 + }, + { + "epoch": 32.93279022403259, + "grad_norm": 0.10465177148580551, + "learning_rate": 2.6344968181871097e-07, + "loss": 0.0104, + "step": 113190 + }, + { + "epoch": 32.93569973814373, + "grad_norm": 0.07207321375608444, + "learning_rate": 2.626789133003993e-07, + "loss": 0.0106, + "step": 113200 + }, + { + "epoch": 32.93860925225487, + "grad_norm": 0.1190493106842041, + "learning_rate": 2.619092639843651e-07, + "loss": 0.0102, + "step": 113210 + }, + { + "epoch": 32.941518766366016, + "grad_norm": 0.10432503372430801, + "learning_rate": 2.6114073392906166e-07, + "loss": 0.0107, + "step": 113220 + }, + { + "epoch": 32.94442828047716, + "grad_norm": 0.10363354533910751, + "learning_rate": 2.603733231928507e-07, + "loss": 0.0081, + "step": 113230 + }, + { + "epoch": 32.94733779458831, + "grad_norm": 0.10553332418203354, + "learning_rate": 2.5960703183401215e-07, + "loss": 0.0105, + "step": 113240 + }, + { + "epoch": 32.950247308699446, + "grad_norm": 0.10801335424184799, + "learning_rate": 2.588418599107428e-07, + "loss": 0.0094, + "step": 113250 + }, + { + "epoch": 32.95315682281059, + "grad_norm": 0.1387072503566742, + "learning_rate": 2.5807780748114963e-07, + "loss": 0.0088, + "step": 113260 + }, + { + "epoch": 32.95606633692174, + "grad_norm": 0.07610314339399338, + "learning_rate": 2.573148746032594e-07, + "loss": 0.0099, + "step": 113270 + }, + { + "epoch": 32.958975851032875, + "grad_norm": 0.1130502000451088, + "learning_rate": 2.565530613350109e-07, + "loss": 0.0104, + "step": 113280 + }, + { + "epoch": 32.96188536514402, + "grad_norm": 0.06077751889824867, + "learning_rate": 2.557923677342594e-07, + "loss": 0.01, + "step": 113290 + }, + { + "epoch": 32.964794879255166, + "grad_norm": 0.06468594819307327, + "learning_rate": 2.550327938587754e-07, + "loss": 0.0087, + "step": 113300 + }, + { + "epoch": 32.967704393366304, + "grad_norm": 0.12071669846773148, + "learning_rate": 2.542743397662395e-07, + "loss": 0.0095, + "step": 113310 + }, + { + "epoch": 32.97061390747745, + "grad_norm": 0.0885557234287262, + "learning_rate": 2.535170055142555e-07, + "loss": 0.009, + "step": 113320 + }, + { + "epoch": 32.973523421588595, + "grad_norm": 0.1973399817943573, + "learning_rate": 2.5276079116033747e-07, + "loss": 0.0129, + "step": 113330 + }, + { + "epoch": 32.97643293569974, + "grad_norm": 0.08399847149848938, + "learning_rate": 2.5200569676191444e-07, + "loss": 0.0098, + "step": 113340 + }, + { + "epoch": 32.97934244981088, + "grad_norm": 0.11781756579875946, + "learning_rate": 2.512517223763272e-07, + "loss": 0.0115, + "step": 113350 + }, + { + "epoch": 32.982251963922025, + "grad_norm": 0.08295045793056488, + "learning_rate": 2.5049886806084163e-07, + "loss": 0.0096, + "step": 113360 + }, + { + "epoch": 32.98516147803317, + "grad_norm": 0.14667880535125732, + "learning_rate": 2.497471338726287e-07, + "loss": 0.0111, + "step": 113370 + }, + { + "epoch": 32.98807099214431, + "grad_norm": 0.07683733105659485, + "learning_rate": 2.48996519868776e-07, + "loss": 0.0118, + "step": 113380 + }, + { + "epoch": 32.990980506255454, + "grad_norm": 0.1357795000076294, + "learning_rate": 2.482470261062897e-07, + "loss": 0.0106, + "step": 113390 + }, + { + "epoch": 32.9938900203666, + "grad_norm": 0.1141398698091507, + "learning_rate": 2.4749865264209083e-07, + "loss": 0.0102, + "step": 113400 + }, + { + "epoch": 32.996799534477745, + "grad_norm": 0.12796789407730103, + "learning_rate": 2.467513995330106e-07, + "loss": 0.0127, + "step": 113410 + }, + { + "epoch": 32.99970904858888, + "grad_norm": 0.11440057307481766, + "learning_rate": 2.4600526683579705e-07, + "loss": 0.0107, + "step": 113420 + }, + { + "epoch": 33.00261856270003, + "grad_norm": 0.11050476878881454, + "learning_rate": 2.4526025460711476e-07, + "loss": 0.0109, + "step": 113430 + }, + { + "epoch": 33.005528076811174, + "grad_norm": 0.08690178394317627, + "learning_rate": 2.4451636290354683e-07, + "loss": 0.0132, + "step": 113440 + }, + { + "epoch": 33.00843759092231, + "grad_norm": 0.057297322899103165, + "learning_rate": 2.4377359178157977e-07, + "loss": 0.009, + "step": 113450 + }, + { + "epoch": 33.01134710503346, + "grad_norm": 0.15307940542697906, + "learning_rate": 2.4303194129762676e-07, + "loss": 0.0095, + "step": 113460 + }, + { + "epoch": 33.014256619144604, + "grad_norm": 0.08126943558454514, + "learning_rate": 2.422914115080127e-07, + "loss": 0.0091, + "step": 113470 + }, + { + "epoch": 33.01716613325575, + "grad_norm": 0.08572998642921448, + "learning_rate": 2.41552002468971e-07, + "loss": 0.0092, + "step": 113480 + }, + { + "epoch": 33.02007564736689, + "grad_norm": 0.12511026859283447, + "learning_rate": 2.4081371423665843e-07, + "loss": 0.0118, + "step": 113490 + }, + { + "epoch": 33.02298516147803, + "grad_norm": 0.1469980776309967, + "learning_rate": 2.4007654686714175e-07, + "loss": 0.0105, + "step": 113500 + }, + { + "epoch": 33.02589467558918, + "grad_norm": 0.09868685156106949, + "learning_rate": 2.3934050041640456e-07, + "loss": 0.0109, + "step": 113510 + }, + { + "epoch": 33.02880418970032, + "grad_norm": 0.09058651328086853, + "learning_rate": 2.386055749403454e-07, + "loss": 0.0125, + "step": 113520 + }, + { + "epoch": 33.03171370381146, + "grad_norm": 0.11022749543190002, + "learning_rate": 2.3787177049477472e-07, + "loss": 0.0097, + "step": 113530 + }, + { + "epoch": 33.03462321792261, + "grad_norm": 0.08677084743976593, + "learning_rate": 2.371390871354212e-07, + "loss": 0.01, + "step": 113540 + }, + { + "epoch": 33.03753273203375, + "grad_norm": 0.09893183410167694, + "learning_rate": 2.364075249179287e-07, + "loss": 0.0117, + "step": 113550 + }, + { + "epoch": 33.04044224614489, + "grad_norm": 0.15411795675754547, + "learning_rate": 2.3567708389785282e-07, + "loss": 0.0097, + "step": 113560 + }, + { + "epoch": 33.04335176025604, + "grad_norm": 0.1190224140882492, + "learning_rate": 2.3494776413066744e-07, + "loss": 0.0115, + "step": 113570 + }, + { + "epoch": 33.04626127436718, + "grad_norm": 0.1034713014960289, + "learning_rate": 2.3421956567175663e-07, + "loss": 0.0079, + "step": 113580 + }, + { + "epoch": 33.04917078847832, + "grad_norm": 0.11078616976737976, + "learning_rate": 2.3349248857642612e-07, + "loss": 0.01, + "step": 113590 + }, + { + "epoch": 33.05208030258947, + "grad_norm": 0.09900974482297897, + "learning_rate": 2.3276653289989014e-07, + "loss": 0.0122, + "step": 113600 + }, + { + "epoch": 33.05498981670061, + "grad_norm": 0.09511622041463852, + "learning_rate": 2.3204169869727953e-07, + "loss": 0.0138, + "step": 113610 + }, + { + "epoch": 33.05789933081176, + "grad_norm": 0.08155988156795502, + "learning_rate": 2.3131798602364197e-07, + "loss": 0.0101, + "step": 113620 + }, + { + "epoch": 33.060808844922896, + "grad_norm": 0.1755489557981491, + "learning_rate": 2.305953949339401e-07, + "loss": 0.012, + "step": 113630 + }, + { + "epoch": 33.06371835903404, + "grad_norm": 0.07273174077272415, + "learning_rate": 2.2987392548304841e-07, + "loss": 0.0109, + "step": 113640 + }, + { + "epoch": 33.06662787314519, + "grad_norm": 0.1844385266304016, + "learning_rate": 2.2915357772575473e-07, + "loss": 0.0105, + "step": 113650 + }, + { + "epoch": 33.069537387256325, + "grad_norm": 0.06846442818641663, + "learning_rate": 2.2843435171677029e-07, + "loss": 0.0099, + "step": 113660 + }, + { + "epoch": 33.07244690136747, + "grad_norm": 0.10699552297592163, + "learning_rate": 2.2771624751071308e-07, + "loss": 0.0104, + "step": 113670 + }, + { + "epoch": 33.075356415478616, + "grad_norm": 0.1048256903886795, + "learning_rate": 2.2699926516211778e-07, + "loss": 0.0128, + "step": 113680 + }, + { + "epoch": 33.07826592958976, + "grad_norm": 0.22278115153312683, + "learning_rate": 2.262834047254342e-07, + "loss": 0.0103, + "step": 113690 + }, + { + "epoch": 33.0811754437009, + "grad_norm": 0.0973890870809555, + "learning_rate": 2.2556866625502881e-07, + "loss": 0.0109, + "step": 113700 + }, + { + "epoch": 33.084084957812046, + "grad_norm": 0.12602873146533966, + "learning_rate": 2.2485504980518156e-07, + "loss": 0.0097, + "step": 113710 + }, + { + "epoch": 33.08699447192319, + "grad_norm": 0.11943813413381577, + "learning_rate": 2.241425554300841e-07, + "loss": 0.0106, + "step": 113720 + }, + { + "epoch": 33.08990398603433, + "grad_norm": 0.07241042703390121, + "learning_rate": 2.2343118318384648e-07, + "loss": 0.0105, + "step": 113730 + }, + { + "epoch": 33.092813500145475, + "grad_norm": 0.09410500526428223, + "learning_rate": 2.2272093312049546e-07, + "loss": 0.0106, + "step": 113740 + }, + { + "epoch": 33.09572301425662, + "grad_norm": 0.11516346037387848, + "learning_rate": 2.2201180529396792e-07, + "loss": 0.0081, + "step": 113750 + }, + { + "epoch": 33.098632528367766, + "grad_norm": 0.0879523828625679, + "learning_rate": 2.2130379975811576e-07, + "loss": 0.0102, + "step": 113760 + }, + { + "epoch": 33.101542042478904, + "grad_norm": 0.10143303871154785, + "learning_rate": 2.20596916566711e-07, + "loss": 0.0114, + "step": 113770 + }, + { + "epoch": 33.10445155659005, + "grad_norm": 0.17294614017009735, + "learning_rate": 2.1989115577343566e-07, + "loss": 0.0118, + "step": 113780 + }, + { + "epoch": 33.107361070701195, + "grad_norm": 0.13975852727890015, + "learning_rate": 2.1918651743188522e-07, + "loss": 0.0114, + "step": 113790 + }, + { + "epoch": 33.11027058481233, + "grad_norm": 0.07805381715297699, + "learning_rate": 2.1848300159557355e-07, + "loss": 0.0106, + "step": 113800 + }, + { + "epoch": 33.11318009892348, + "grad_norm": 0.11508242040872574, + "learning_rate": 2.1778060831792957e-07, + "loss": 0.0096, + "step": 113810 + }, + { + "epoch": 33.116089613034625, + "grad_norm": 0.1194833368062973, + "learning_rate": 2.1707933765229392e-07, + "loss": 0.0125, + "step": 113820 + }, + { + "epoch": 33.11899912714577, + "grad_norm": 0.12383440881967545, + "learning_rate": 2.1637918965192405e-07, + "loss": 0.0094, + "step": 113830 + }, + { + "epoch": 33.12190864125691, + "grad_norm": 0.06998740136623383, + "learning_rate": 2.1568016436998904e-07, + "loss": 0.0106, + "step": 113840 + }, + { + "epoch": 33.124818155368054, + "grad_norm": 0.1258663535118103, + "learning_rate": 2.1498226185957981e-07, + "loss": 0.0086, + "step": 113850 + }, + { + "epoch": 33.1277276694792, + "grad_norm": 0.07482526451349258, + "learning_rate": 2.1428548217369558e-07, + "loss": 0.009, + "step": 113860 + }, + { + "epoch": 33.13063718359034, + "grad_norm": 0.10394585132598877, + "learning_rate": 2.1358982536524907e-07, + "loss": 0.009, + "step": 113870 + }, + { + "epoch": 33.13354669770148, + "grad_norm": 0.07794881612062454, + "learning_rate": 2.1289529148707632e-07, + "loss": 0.0103, + "step": 113880 + }, + { + "epoch": 33.13645621181263, + "grad_norm": 0.09218138456344604, + "learning_rate": 2.1220188059191848e-07, + "loss": 0.0092, + "step": 113890 + }, + { + "epoch": 33.139365725923774, + "grad_norm": 0.07111234217882156, + "learning_rate": 2.1150959273243674e-07, + "loss": 0.0108, + "step": 113900 + }, + { + "epoch": 33.14227524003491, + "grad_norm": 0.09048547595739365, + "learning_rate": 2.1081842796120575e-07, + "loss": 0.0119, + "step": 113910 + }, + { + "epoch": 33.14518475414606, + "grad_norm": 0.09072431176900864, + "learning_rate": 2.1012838633071518e-07, + "loss": 0.0112, + "step": 113920 + }, + { + "epoch": 33.1480942682572, + "grad_norm": 0.08520019799470901, + "learning_rate": 2.0943946789336975e-07, + "loss": 0.0102, + "step": 113930 + }, + { + "epoch": 33.15100378236834, + "grad_norm": 0.11448948085308075, + "learning_rate": 2.087516727014893e-07, + "loss": 0.0103, + "step": 113940 + }, + { + "epoch": 33.15391329647949, + "grad_norm": 0.12774047255516052, + "learning_rate": 2.080650008073004e-07, + "loss": 0.012, + "step": 113950 + }, + { + "epoch": 33.15682281059063, + "grad_norm": 0.21321485936641693, + "learning_rate": 2.073794522629613e-07, + "loss": 0.009, + "step": 113960 + }, + { + "epoch": 33.15973232470178, + "grad_norm": 0.10949594527482986, + "learning_rate": 2.066950271205287e-07, + "loss": 0.0106, + "step": 113970 + }, + { + "epoch": 33.16264183881292, + "grad_norm": 0.08765363693237305, + "learning_rate": 2.0601172543197933e-07, + "loss": 0.0123, + "step": 113980 + }, + { + "epoch": 33.16555135292406, + "grad_norm": 0.16529372334480286, + "learning_rate": 2.053295472492084e-07, + "loss": 0.0097, + "step": 113990 + }, + { + "epoch": 33.16846086703521, + "grad_norm": 0.11635785549879074, + "learning_rate": 2.046484926240244e-07, + "loss": 0.0096, + "step": 114000 + }, + { + "epoch": 33.171370381146346, + "grad_norm": 0.10841956734657288, + "learning_rate": 2.0396856160814603e-07, + "loss": 0.0103, + "step": 114010 + }, + { + "epoch": 33.17427989525749, + "grad_norm": 0.09105563908815384, + "learning_rate": 2.0328975425320694e-07, + "loss": 0.0108, + "step": 114020 + }, + { + "epoch": 33.17718940936864, + "grad_norm": 0.0722188949584961, + "learning_rate": 2.0261207061076425e-07, + "loss": 0.0104, + "step": 114030 + }, + { + "epoch": 33.180098923479775, + "grad_norm": 0.08599385619163513, + "learning_rate": 2.019355107322801e-07, + "loss": 0.0104, + "step": 114040 + }, + { + "epoch": 33.18300843759092, + "grad_norm": 0.1578143835067749, + "learning_rate": 2.0126007466913344e-07, + "loss": 0.0104, + "step": 114050 + }, + { + "epoch": 33.185917951702066, + "grad_norm": 0.09752547740936279, + "learning_rate": 2.005857624726215e-07, + "loss": 0.0132, + "step": 114060 + }, + { + "epoch": 33.18882746581321, + "grad_norm": 0.07992703467607498, + "learning_rate": 1.9991257419395502e-07, + "loss": 0.0102, + "step": 114070 + }, + { + "epoch": 33.19173697992435, + "grad_norm": 0.10856842994689941, + "learning_rate": 1.992405098842548e-07, + "loss": 0.0112, + "step": 114080 + }, + { + "epoch": 33.194646494035496, + "grad_norm": 0.09614811092615128, + "learning_rate": 1.9856956959456163e-07, + "loss": 0.0104, + "step": 114090 + }, + { + "epoch": 33.19755600814664, + "grad_norm": 0.1142117828130722, + "learning_rate": 1.9789975337582812e-07, + "loss": 0.0106, + "step": 114100 + }, + { + "epoch": 33.20046552225778, + "grad_norm": 0.13116233050823212, + "learning_rate": 1.972310612789252e-07, + "loss": 0.0105, + "step": 114110 + }, + { + "epoch": 33.203375036368925, + "grad_norm": 0.12359324842691422, + "learning_rate": 1.965634933546323e-07, + "loss": 0.0114, + "step": 114120 + }, + { + "epoch": 33.20628455048007, + "grad_norm": 0.10956642031669617, + "learning_rate": 1.958970496536472e-07, + "loss": 0.0107, + "step": 114130 + }, + { + "epoch": 33.209194064591216, + "grad_norm": 0.06521038711071014, + "learning_rate": 1.952317302265827e-07, + "loss": 0.01, + "step": 114140 + }, + { + "epoch": 33.212103578702354, + "grad_norm": 0.14233016967773438, + "learning_rate": 1.9456753512396508e-07, + "loss": 0.0086, + "step": 114150 + }, + { + "epoch": 33.2150130928135, + "grad_norm": 0.09640666842460632, + "learning_rate": 1.9390446439623733e-07, + "loss": 0.0111, + "step": 114160 + }, + { + "epoch": 33.217922606924645, + "grad_norm": 0.11295308917760849, + "learning_rate": 1.932425180937525e-07, + "loss": 0.0124, + "step": 114170 + }, + { + "epoch": 33.220832121035784, + "grad_norm": 0.09635929763317108, + "learning_rate": 1.92581696266782e-07, + "loss": 0.009, + "step": 114180 + }, + { + "epoch": 33.22374163514693, + "grad_norm": 0.07521814107894897, + "learning_rate": 1.9192199896551243e-07, + "loss": 0.0104, + "step": 114190 + }, + { + "epoch": 33.226651149258075, + "grad_norm": 0.09553618729114532, + "learning_rate": 1.91263426240042e-07, + "loss": 0.0108, + "step": 114200 + }, + { + "epoch": 33.22956066336922, + "grad_norm": 0.12144821137189865, + "learning_rate": 1.9060597814038404e-07, + "loss": 0.0103, + "step": 114210 + }, + { + "epoch": 33.23247017748036, + "grad_norm": 0.15452967584133148, + "learning_rate": 1.8994965471646863e-07, + "loss": 0.0105, + "step": 114220 + }, + { + "epoch": 33.235379691591504, + "grad_norm": 0.10495667904615402, + "learning_rate": 1.892944560181409e-07, + "loss": 0.0105, + "step": 114230 + }, + { + "epoch": 33.23828920570265, + "grad_norm": 0.11087114363908768, + "learning_rate": 1.8864038209515598e-07, + "loss": 0.0098, + "step": 114240 + }, + { + "epoch": 33.24119871981379, + "grad_norm": 0.12453994899988174, + "learning_rate": 1.8798743299718757e-07, + "loss": 0.0108, + "step": 114250 + }, + { + "epoch": 33.24410823392493, + "grad_norm": 0.07090442627668381, + "learning_rate": 1.873356087738226e-07, + "loss": 0.0084, + "step": 114260 + }, + { + "epoch": 33.24701774803608, + "grad_norm": 0.06383980065584183, + "learning_rate": 1.8668490947456317e-07, + "loss": 0.0102, + "step": 114270 + }, + { + "epoch": 33.249927262147224, + "grad_norm": 0.10614398121833801, + "learning_rate": 1.8603533514882642e-07, + "loss": 0.0109, + "step": 114280 + }, + { + "epoch": 33.25283677625836, + "grad_norm": 0.17299754917621613, + "learning_rate": 1.8538688584594123e-07, + "loss": 0.0097, + "step": 114290 + }, + { + "epoch": 33.25574629036951, + "grad_norm": 0.08757849782705307, + "learning_rate": 1.8473956161515481e-07, + "loss": 0.0098, + "step": 114300 + }, + { + "epoch": 33.258655804480654, + "grad_norm": 0.14491645991802216, + "learning_rate": 1.8409336250562626e-07, + "loss": 0.0123, + "step": 114310 + }, + { + "epoch": 33.26156531859179, + "grad_norm": 0.0942610651254654, + "learning_rate": 1.834482885664296e-07, + "loss": 0.0097, + "step": 114320 + }, + { + "epoch": 33.26447483270294, + "grad_norm": 0.15184688568115234, + "learning_rate": 1.82804339846554e-07, + "loss": 0.0113, + "step": 114330 + }, + { + "epoch": 33.26738434681408, + "grad_norm": 0.08741419017314911, + "learning_rate": 1.8216151639490696e-07, + "loss": 0.0091, + "step": 114340 + }, + { + "epoch": 33.27029386092523, + "grad_norm": 0.09582412242889404, + "learning_rate": 1.8151981826030117e-07, + "loss": 0.0093, + "step": 114350 + }, + { + "epoch": 33.27320337503637, + "grad_norm": 0.07632077485322952, + "learning_rate": 1.8087924549147094e-07, + "loss": 0.0084, + "step": 114360 + }, + { + "epoch": 33.27611288914751, + "grad_norm": 0.12936431169509888, + "learning_rate": 1.802397981370657e-07, + "loss": 0.0113, + "step": 114370 + }, + { + "epoch": 33.27902240325866, + "grad_norm": 0.0838799923658371, + "learning_rate": 1.796014762456466e-07, + "loss": 0.01, + "step": 114380 + }, + { + "epoch": 33.281931917369796, + "grad_norm": 0.12660086154937744, + "learning_rate": 1.789642798656882e-07, + "loss": 0.0112, + "step": 114390 + }, + { + "epoch": 33.28484143148094, + "grad_norm": 0.09279454499483109, + "learning_rate": 1.7832820904558178e-07, + "loss": 0.0103, + "step": 114400 + }, + { + "epoch": 33.28775094559209, + "grad_norm": 0.10074537992477417, + "learning_rate": 1.7769326383363538e-07, + "loss": 0.0122, + "step": 114410 + }, + { + "epoch": 33.29066045970323, + "grad_norm": 0.13696125149726868, + "learning_rate": 1.7705944427806543e-07, + "loss": 0.0098, + "step": 114420 + }, + { + "epoch": 33.29356997381437, + "grad_norm": 0.1050359383225441, + "learning_rate": 1.764267504270084e-07, + "loss": 0.0106, + "step": 114430 + }, + { + "epoch": 33.29647948792552, + "grad_norm": 0.17575575411319733, + "learning_rate": 1.7579518232851256e-07, + "loss": 0.0108, + "step": 114440 + }, + { + "epoch": 33.29938900203666, + "grad_norm": 0.09463081508874893, + "learning_rate": 1.7516474003054118e-07, + "loss": 0.0127, + "step": 114450 + }, + { + "epoch": 33.3022985161478, + "grad_norm": 0.12295771390199661, + "learning_rate": 1.7453542358097097e-07, + "loss": 0.0094, + "step": 114460 + }, + { + "epoch": 33.305208030258946, + "grad_norm": 0.08456528931856155, + "learning_rate": 1.7390723302759704e-07, + "loss": 0.0107, + "step": 114470 + }, + { + "epoch": 33.30811754437009, + "grad_norm": 0.10078727453947067, + "learning_rate": 1.7328016841812622e-07, + "loss": 0.0111, + "step": 114480 + }, + { + "epoch": 33.31102705848124, + "grad_norm": 0.09858165681362152, + "learning_rate": 1.7265422980017709e-07, + "loss": 0.0095, + "step": 114490 + }, + { + "epoch": 33.313936572592375, + "grad_norm": 0.09099265933036804, + "learning_rate": 1.720294172212883e-07, + "loss": 0.011, + "step": 114500 + }, + { + "epoch": 33.31684608670352, + "grad_norm": 0.10051220655441284, + "learning_rate": 1.7140573072890853e-07, + "loss": 0.0088, + "step": 114510 + }, + { + "epoch": 33.319755600814666, + "grad_norm": 0.13587065041065216, + "learning_rate": 1.7078317037040326e-07, + "loss": 0.0135, + "step": 114520 + }, + { + "epoch": 33.322665114925805, + "grad_norm": 0.09997562319040298, + "learning_rate": 1.7016173619305298e-07, + "loss": 0.011, + "step": 114530 + }, + { + "epoch": 33.32557462903695, + "grad_norm": 0.09763843566179276, + "learning_rate": 1.6954142824404995e-07, + "loss": 0.0109, + "step": 114540 + }, + { + "epoch": 33.328484143148096, + "grad_norm": 0.16648532450199127, + "learning_rate": 1.689222465705015e-07, + "loss": 0.0093, + "step": 114550 + }, + { + "epoch": 33.33139365725924, + "grad_norm": 0.15742386877536774, + "learning_rate": 1.683041912194333e-07, + "loss": 0.0128, + "step": 114560 + }, + { + "epoch": 33.33430317137038, + "grad_norm": 0.1359497457742691, + "learning_rate": 1.676872622377812e-07, + "loss": 0.0113, + "step": 114570 + }, + { + "epoch": 33.337212685481525, + "grad_norm": 0.20723038911819458, + "learning_rate": 1.67071459672396e-07, + "loss": 0.0098, + "step": 114580 + }, + { + "epoch": 33.34012219959267, + "grad_norm": 0.12297632545232773, + "learning_rate": 1.6645678357004367e-07, + "loss": 0.0117, + "step": 114590 + }, + { + "epoch": 33.34303171370381, + "grad_norm": 0.0676383301615715, + "learning_rate": 1.6584323397740853e-07, + "loss": 0.0104, + "step": 114600 + }, + { + "epoch": 33.345941227814954, + "grad_norm": 0.17722873389720917, + "learning_rate": 1.6523081094108161e-07, + "loss": 0.011, + "step": 114610 + }, + { + "epoch": 33.3488507419261, + "grad_norm": 0.07686855643987656, + "learning_rate": 1.646195145075724e-07, + "loss": 0.0083, + "step": 114620 + }, + { + "epoch": 33.351760256037245, + "grad_norm": 0.07596100121736526, + "learning_rate": 1.6400934472330708e-07, + "loss": 0.0111, + "step": 114630 + }, + { + "epoch": 33.354669770148384, + "grad_norm": 0.12391330301761627, + "learning_rate": 1.634003016346236e-07, + "loss": 0.0127, + "step": 114640 + }, + { + "epoch": 33.35757928425953, + "grad_norm": 0.10760276764631271, + "learning_rate": 1.627923852877733e-07, + "loss": 0.0096, + "step": 114650 + }, + { + "epoch": 33.360488798370675, + "grad_norm": 0.1168861985206604, + "learning_rate": 1.6218559572892422e-07, + "loss": 0.0107, + "step": 114660 + }, + { + "epoch": 33.36339831248181, + "grad_norm": 0.10089998692274094, + "learning_rate": 1.6157993300415786e-07, + "loss": 0.0098, + "step": 114670 + }, + { + "epoch": 33.36630782659296, + "grad_norm": 0.10503072291612625, + "learning_rate": 1.609753971594724e-07, + "loss": 0.0117, + "step": 114680 + }, + { + "epoch": 33.369217340704104, + "grad_norm": 0.11273769289255142, + "learning_rate": 1.6037198824077447e-07, + "loss": 0.0125, + "step": 114690 + }, + { + "epoch": 33.37212685481525, + "grad_norm": 0.1396297961473465, + "learning_rate": 1.5976970629389076e-07, + "loss": 0.0118, + "step": 114700 + }, + { + "epoch": 33.37503636892639, + "grad_norm": 0.10560968518257141, + "learning_rate": 1.5916855136456298e-07, + "loss": 0.011, + "step": 114710 + }, + { + "epoch": 33.37794588303753, + "grad_norm": 0.09577468037605286, + "learning_rate": 1.5856852349844131e-07, + "loss": 0.0089, + "step": 114720 + }, + { + "epoch": 33.38085539714868, + "grad_norm": 0.14094744622707367, + "learning_rate": 1.5796962274109595e-07, + "loss": 0.0095, + "step": 114730 + }, + { + "epoch": 33.38376491125982, + "grad_norm": 0.12618467211723328, + "learning_rate": 1.5737184913800718e-07, + "loss": 0.011, + "step": 114740 + }, + { + "epoch": 33.38667442537096, + "grad_norm": 0.11381573975086212, + "learning_rate": 1.5677520273457534e-07, + "loss": 0.0117, + "step": 114750 + }, + { + "epoch": 33.38958393948211, + "grad_norm": 0.07933652400970459, + "learning_rate": 1.5617968357610922e-07, + "loss": 0.0108, + "step": 114760 + }, + { + "epoch": 33.39249345359325, + "grad_norm": 0.1643393486738205, + "learning_rate": 1.5558529170783597e-07, + "loss": 0.0123, + "step": 114770 + }, + { + "epoch": 33.39540296770439, + "grad_norm": 0.11940521746873856, + "learning_rate": 1.5499202717489614e-07, + "loss": 0.0117, + "step": 114780 + }, + { + "epoch": 33.39831248181554, + "grad_norm": 0.14546485245227814, + "learning_rate": 1.54399890022342e-07, + "loss": 0.0115, + "step": 114790 + }, + { + "epoch": 33.40122199592668, + "grad_norm": 0.07253575325012207, + "learning_rate": 1.5380888029514428e-07, + "loss": 0.0107, + "step": 114800 + }, + { + "epoch": 33.40413151003782, + "grad_norm": 0.0755431205034256, + "learning_rate": 1.532189980381854e-07, + "loss": 0.011, + "step": 114810 + }, + { + "epoch": 33.40704102414897, + "grad_norm": 0.09480338543653488, + "learning_rate": 1.526302432962645e-07, + "loss": 0.009, + "step": 114820 + }, + { + "epoch": 33.40995053826011, + "grad_norm": 0.08850789815187454, + "learning_rate": 1.5204261611409088e-07, + "loss": 0.0101, + "step": 114830 + }, + { + "epoch": 33.41286005237125, + "grad_norm": 0.08058708161115646, + "learning_rate": 1.5145611653629378e-07, + "loss": 0.0098, + "step": 114840 + }, + { + "epoch": 33.415769566482396, + "grad_norm": 0.08390560001134872, + "learning_rate": 1.5087074460741257e-07, + "loss": 0.009, + "step": 114850 + }, + { + "epoch": 33.41867908059354, + "grad_norm": 0.09424861520528793, + "learning_rate": 1.5028650037190171e-07, + "loss": 0.0093, + "step": 114860 + }, + { + "epoch": 33.42158859470469, + "grad_norm": 0.10516995191574097, + "learning_rate": 1.497033838741324e-07, + "loss": 0.011, + "step": 114870 + }, + { + "epoch": 33.424498108815826, + "grad_norm": 0.1237846314907074, + "learning_rate": 1.4912139515838752e-07, + "loss": 0.011, + "step": 114880 + }, + { + "epoch": 33.42740762292697, + "grad_norm": 0.11270633339881897, + "learning_rate": 1.4854053426886505e-07, + "loss": 0.0092, + "step": 114890 + }, + { + "epoch": 33.43031713703812, + "grad_norm": 0.14799898862838745, + "learning_rate": 1.4796080124967803e-07, + "loss": 0.0118, + "step": 114900 + }, + { + "epoch": 33.433226651149255, + "grad_norm": 0.061912670731544495, + "learning_rate": 1.473821961448546e-07, + "loss": 0.0095, + "step": 114910 + }, + { + "epoch": 33.4361361652604, + "grad_norm": 0.09885569661855698, + "learning_rate": 1.4680471899833292e-07, + "loss": 0.009, + "step": 114920 + }, + { + "epoch": 33.439045679371546, + "grad_norm": 0.10387212038040161, + "learning_rate": 1.4622836985396958e-07, + "loss": 0.0109, + "step": 114930 + }, + { + "epoch": 33.44195519348269, + "grad_norm": 0.10074307024478912, + "learning_rate": 1.4565314875553626e-07, + "loss": 0.0108, + "step": 114940 + }, + { + "epoch": 33.44486470759383, + "grad_norm": 0.11219041794538498, + "learning_rate": 1.4507905574671797e-07, + "loss": 0.0112, + "step": 114950 + }, + { + "epoch": 33.447774221704975, + "grad_norm": 0.08163093775510788, + "learning_rate": 1.445060908711082e-07, + "loss": 0.0093, + "step": 114960 + }, + { + "epoch": 33.45068373581612, + "grad_norm": 0.19116206467151642, + "learning_rate": 1.439342541722255e-07, + "loss": 0.0089, + "step": 114970 + }, + { + "epoch": 33.45359324992726, + "grad_norm": 0.10185519605875015, + "learning_rate": 1.433635456934934e-07, + "loss": 0.0099, + "step": 114980 + }, + { + "epoch": 33.456502764038405, + "grad_norm": 0.12924736738204956, + "learning_rate": 1.4279396547825564e-07, + "loss": 0.012, + "step": 114990 + }, + { + "epoch": 33.45941227814955, + "grad_norm": 0.0821365937590599, + "learning_rate": 1.4222551356976587e-07, + "loss": 0.0112, + "step": 115000 + }, + { + "epoch": 33.462321792260695, + "grad_norm": 0.12972381711006165, + "learning_rate": 1.4165819001119628e-07, + "loss": 0.0093, + "step": 115010 + }, + { + "epoch": 33.465231306371834, + "grad_norm": 0.11479459702968597, + "learning_rate": 1.4109199484563075e-07, + "loss": 0.0095, + "step": 115020 + }, + { + "epoch": 33.46814082048298, + "grad_norm": 0.09311886131763458, + "learning_rate": 1.4052692811606816e-07, + "loss": 0.0111, + "step": 115030 + }, + { + "epoch": 33.471050334594125, + "grad_norm": 0.13909149169921875, + "learning_rate": 1.3996298986542087e-07, + "loss": 0.0095, + "step": 115040 + }, + { + "epoch": 33.47395984870526, + "grad_norm": 0.09061503410339355, + "learning_rate": 1.39400180136518e-07, + "loss": 0.0126, + "step": 115050 + }, + { + "epoch": 33.47686936281641, + "grad_norm": 0.10700366646051407, + "learning_rate": 1.3883849897209865e-07, + "loss": 0.0105, + "step": 115060 + }, + { + "epoch": 33.479778876927554, + "grad_norm": 0.1023109182715416, + "learning_rate": 1.382779464148204e-07, + "loss": 0.0112, + "step": 115070 + }, + { + "epoch": 33.4826883910387, + "grad_norm": 0.20828597247600555, + "learning_rate": 1.3771852250725413e-07, + "loss": 0.0111, + "step": 115080 + }, + { + "epoch": 33.48559790514984, + "grad_norm": 0.08630622178316116, + "learning_rate": 1.371602272918826e-07, + "loss": 0.0098, + "step": 115090 + }, + { + "epoch": 33.48850741926098, + "grad_norm": 0.06674717366695404, + "learning_rate": 1.3660306081110517e-07, + "loss": 0.0114, + "step": 115100 + }, + { + "epoch": 33.49141693337213, + "grad_norm": 0.09136655926704407, + "learning_rate": 1.360470231072364e-07, + "loss": 0.01, + "step": 115110 + }, + { + "epoch": 33.49432644748327, + "grad_norm": 0.1288871169090271, + "learning_rate": 1.354921142225024e-07, + "loss": 0.0096, + "step": 115120 + }, + { + "epoch": 33.49723596159441, + "grad_norm": 0.08691372722387314, + "learning_rate": 1.3493833419904454e-07, + "loss": 0.0094, + "step": 115130 + }, + { + "epoch": 33.50014547570556, + "grad_norm": 0.11058409512042999, + "learning_rate": 1.3438568307892085e-07, + "loss": 0.0112, + "step": 115140 + }, + { + "epoch": 33.503054989816704, + "grad_norm": 0.18543240427970886, + "learning_rate": 1.338341609040994e-07, + "loss": 0.0101, + "step": 115150 + }, + { + "epoch": 33.50596450392784, + "grad_norm": 0.1164708137512207, + "learning_rate": 1.3328376771646334e-07, + "loss": 0.0088, + "step": 115160 + }, + { + "epoch": 33.50887401803899, + "grad_norm": 0.07711146026849747, + "learning_rate": 1.327345035578159e-07, + "loss": 0.0097, + "step": 115170 + }, + { + "epoch": 33.51178353215013, + "grad_norm": 0.11332345753908157, + "learning_rate": 1.3218636846986543e-07, + "loss": 0.0119, + "step": 115180 + }, + { + "epoch": 33.51469304626127, + "grad_norm": 0.20833554863929749, + "learning_rate": 1.316393624942436e-07, + "loss": 0.0117, + "step": 115190 + }, + { + "epoch": 33.51760256037242, + "grad_norm": 0.10343638807535172, + "learning_rate": 1.3109348567248712e-07, + "loss": 0.011, + "step": 115200 + }, + { + "epoch": 33.52051207448356, + "grad_norm": 0.10770288109779358, + "learning_rate": 1.3054873804605627e-07, + "loss": 0.0082, + "step": 115210 + }, + { + "epoch": 33.52342158859471, + "grad_norm": 0.10210409760475159, + "learning_rate": 1.300051196563179e-07, + "loss": 0.0097, + "step": 115220 + }, + { + "epoch": 33.526331102705846, + "grad_norm": 0.07165789604187012, + "learning_rate": 1.294626305445573e-07, + "loss": 0.0099, + "step": 115230 + }, + { + "epoch": 33.52924061681699, + "grad_norm": 0.09223448485136032, + "learning_rate": 1.2892127075197324e-07, + "loss": 0.0101, + "step": 115240 + }, + { + "epoch": 33.53215013092814, + "grad_norm": 0.17054957151412964, + "learning_rate": 1.2838104031967778e-07, + "loss": 0.0126, + "step": 115250 + }, + { + "epoch": 33.535059645039276, + "grad_norm": 0.09729977697134018, + "learning_rate": 1.2784193928869647e-07, + "loss": 0.01, + "step": 115260 + }, + { + "epoch": 33.53796915915042, + "grad_norm": 0.24766743183135986, + "learning_rate": 1.2730396769997487e-07, + "loss": 0.0102, + "step": 115270 + }, + { + "epoch": 33.54087867326157, + "grad_norm": 0.10492187738418579, + "learning_rate": 1.267671255943653e-07, + "loss": 0.0109, + "step": 115280 + }, + { + "epoch": 33.54378818737271, + "grad_norm": 0.12841401994228363, + "learning_rate": 1.2623141301263686e-07, + "loss": 0.0107, + "step": 115290 + }, + { + "epoch": 33.54669770148385, + "grad_norm": 0.06626275926828384, + "learning_rate": 1.256968299954736e-07, + "loss": 0.0074, + "step": 115300 + }, + { + "epoch": 33.549607215594996, + "grad_norm": 0.08722301572561264, + "learning_rate": 1.2516337658347643e-07, + "loss": 0.0098, + "step": 115310 + }, + { + "epoch": 33.55251672970614, + "grad_norm": 0.08827709406614304, + "learning_rate": 1.2463105281715458e-07, + "loss": 0.0093, + "step": 115320 + }, + { + "epoch": 33.55542624381728, + "grad_norm": 0.09700597077608109, + "learning_rate": 1.240998587369341e-07, + "loss": 0.0106, + "step": 115330 + }, + { + "epoch": 33.558335757928425, + "grad_norm": 0.08102884888648987, + "learning_rate": 1.2356979438315764e-07, + "loss": 0.0093, + "step": 115340 + }, + { + "epoch": 33.56124527203957, + "grad_norm": 0.09546265006065369, + "learning_rate": 1.230408597960797e-07, + "loss": 0.0108, + "step": 115350 + }, + { + "epoch": 33.56415478615071, + "grad_norm": 0.0673157274723053, + "learning_rate": 1.2251305501586818e-07, + "loss": 0.0121, + "step": 115360 + }, + { + "epoch": 33.567064300261855, + "grad_norm": 0.11667505651712418, + "learning_rate": 1.2198638008260765e-07, + "loss": 0.0108, + "step": 115370 + }, + { + "epoch": 33.569973814373, + "grad_norm": 0.14977186918258667, + "learning_rate": 1.2146083503629612e-07, + "loss": 0.0115, + "step": 115380 + }, + { + "epoch": 33.572883328484146, + "grad_norm": 0.07893355190753937, + "learning_rate": 1.20936419916845e-07, + "loss": 0.0104, + "step": 115390 + }, + { + "epoch": 33.575792842595284, + "grad_norm": 0.13183772563934326, + "learning_rate": 1.2041313476407744e-07, + "loss": 0.0105, + "step": 115400 + }, + { + "epoch": 33.57870235670643, + "grad_norm": 0.15581096708774567, + "learning_rate": 1.1989097961773666e-07, + "loss": 0.0109, + "step": 115410 + }, + { + "epoch": 33.581611870817575, + "grad_norm": 0.1550799459218979, + "learning_rate": 1.193699545174759e-07, + "loss": 0.0111, + "step": 115420 + }, + { + "epoch": 33.58452138492871, + "grad_norm": 0.08831837773323059, + "learning_rate": 1.1885005950286354e-07, + "loss": 0.011, + "step": 115430 + }, + { + "epoch": 33.58743089903986, + "grad_norm": 0.12038827687501907, + "learning_rate": 1.1833129461338133e-07, + "loss": 0.0104, + "step": 115440 + }, + { + "epoch": 33.590340413151004, + "grad_norm": 0.2220965474843979, + "learning_rate": 1.1781365988842774e-07, + "loss": 0.0096, + "step": 115450 + }, + { + "epoch": 33.59324992726215, + "grad_norm": 0.12044229358434677, + "learning_rate": 1.1729715536731134e-07, + "loss": 0.0113, + "step": 115460 + }, + { + "epoch": 33.59615944137329, + "grad_norm": 0.19002002477645874, + "learning_rate": 1.1678178108926075e-07, + "loss": 0.0114, + "step": 115470 + }, + { + "epoch": 33.599068955484434, + "grad_norm": 0.06616521626710892, + "learning_rate": 1.1626753709341132e-07, + "loss": 0.009, + "step": 115480 + }, + { + "epoch": 33.60197846959558, + "grad_norm": 0.11439889669418335, + "learning_rate": 1.1575442341882014e-07, + "loss": 0.0093, + "step": 115490 + }, + { + "epoch": 33.60488798370672, + "grad_norm": 0.158896803855896, + "learning_rate": 1.1524244010445106e-07, + "loss": 0.0098, + "step": 115500 + }, + { + "epoch": 33.60779749781786, + "grad_norm": 0.10217086970806122, + "learning_rate": 1.1473158718918963e-07, + "loss": 0.0115, + "step": 115510 + }, + { + "epoch": 33.61070701192901, + "grad_norm": 0.10855524241924286, + "learning_rate": 1.1422186471182816e-07, + "loss": 0.0122, + "step": 115520 + }, + { + "epoch": 33.613616526040154, + "grad_norm": 0.128092423081398, + "learning_rate": 1.1371327271107901e-07, + "loss": 0.0116, + "step": 115530 + }, + { + "epoch": 33.61652604015129, + "grad_norm": 0.08798848092556, + "learning_rate": 1.1320581122556628e-07, + "loss": 0.0095, + "step": 115540 + }, + { + "epoch": 33.61943555426244, + "grad_norm": 0.14515765011310577, + "learning_rate": 1.1269948029382915e-07, + "loss": 0.0107, + "step": 115550 + }, + { + "epoch": 33.62234506837358, + "grad_norm": 0.10591159760951996, + "learning_rate": 1.1219427995431685e-07, + "loss": 0.0102, + "step": 115560 + }, + { + "epoch": 33.62525458248472, + "grad_norm": 0.15864084661006927, + "learning_rate": 1.1169021024539872e-07, + "loss": 0.0101, + "step": 115570 + }, + { + "epoch": 33.62816409659587, + "grad_norm": 0.07464896887540817, + "learning_rate": 1.111872712053541e-07, + "loss": 0.0109, + "step": 115580 + }, + { + "epoch": 33.63107361070701, + "grad_norm": 0.08895702660083771, + "learning_rate": 1.1068546287237913e-07, + "loss": 0.0101, + "step": 115590 + }, + { + "epoch": 33.63398312481816, + "grad_norm": 0.15505990386009216, + "learning_rate": 1.1018478528458332e-07, + "loss": 0.0122, + "step": 115600 + }, + { + "epoch": 33.6368926389293, + "grad_norm": 0.14215737581253052, + "learning_rate": 1.096852384799879e-07, + "loss": 0.01, + "step": 115610 + }, + { + "epoch": 33.63980215304044, + "grad_norm": 0.08754459768533707, + "learning_rate": 1.0918682249653255e-07, + "loss": 0.0095, + "step": 115620 + }, + { + "epoch": 33.64271166715159, + "grad_norm": 0.08705062419176102, + "learning_rate": 1.0868953737206533e-07, + "loss": 0.0097, + "step": 115630 + }, + { + "epoch": 33.645621181262726, + "grad_norm": 0.1044706329703331, + "learning_rate": 1.0819338314435267e-07, + "loss": 0.0102, + "step": 115640 + }, + { + "epoch": 33.64853069537387, + "grad_norm": 0.09619296342134476, + "learning_rate": 1.0769835985107779e-07, + "loss": 0.0105, + "step": 115650 + }, + { + "epoch": 33.65144020948502, + "grad_norm": 0.09079606086015701, + "learning_rate": 1.072044675298306e-07, + "loss": 0.0105, + "step": 115660 + }, + { + "epoch": 33.65434972359616, + "grad_norm": 0.08917015790939331, + "learning_rate": 1.0671170621811943e-07, + "loss": 0.0096, + "step": 115670 + }, + { + "epoch": 33.6572592377073, + "grad_norm": 0.10294067859649658, + "learning_rate": 1.0622007595336936e-07, + "loss": 0.0118, + "step": 115680 + }, + { + "epoch": 33.660168751818446, + "grad_norm": 0.08469512313604355, + "learning_rate": 1.0572957677291217e-07, + "loss": 0.0102, + "step": 115690 + }, + { + "epoch": 33.66307826592959, + "grad_norm": 0.0809379369020462, + "learning_rate": 1.0524020871400141e-07, + "loss": 0.0079, + "step": 115700 + }, + { + "epoch": 33.66598778004073, + "grad_norm": 0.1469736546278, + "learning_rate": 1.0475197181379903e-07, + "loss": 0.0111, + "step": 115710 + }, + { + "epoch": 33.668897294151876, + "grad_norm": 0.09385460615158081, + "learning_rate": 1.0426486610938368e-07, + "loss": 0.0097, + "step": 115720 + }, + { + "epoch": 33.67180680826302, + "grad_norm": 0.12486802786588669, + "learning_rate": 1.0377889163774911e-07, + "loss": 0.0104, + "step": 115730 + }, + { + "epoch": 33.67471632237417, + "grad_norm": 0.09400205314159393, + "learning_rate": 1.0329404843580248e-07, + "loss": 0.011, + "step": 115740 + }, + { + "epoch": 33.677625836485305, + "grad_norm": 0.15378637611865997, + "learning_rate": 1.0281033654036098e-07, + "loss": 0.0107, + "step": 115750 + }, + { + "epoch": 33.68053535059645, + "grad_norm": 0.052655190229415894, + "learning_rate": 1.0232775598816358e-07, + "loss": 0.0105, + "step": 115760 + }, + { + "epoch": 33.683444864707596, + "grad_norm": 0.1564498394727707, + "learning_rate": 1.0184630681585427e-07, + "loss": 0.0102, + "step": 115770 + }, + { + "epoch": 33.686354378818734, + "grad_norm": 0.09940314292907715, + "learning_rate": 1.013659890600005e-07, + "loss": 0.0097, + "step": 115780 + }, + { + "epoch": 33.68926389292988, + "grad_norm": 0.0868314728140831, + "learning_rate": 1.0088680275707807e-07, + "loss": 0.0111, + "step": 115790 + }, + { + "epoch": 33.692173407041025, + "grad_norm": 0.2319600135087967, + "learning_rate": 1.004087479434762e-07, + "loss": 0.0139, + "step": 115800 + }, + { + "epoch": 33.69508292115217, + "grad_norm": 0.08892180770635605, + "learning_rate": 9.993182465550254e-08, + "loss": 0.0095, + "step": 115810 + }, + { + "epoch": 33.69799243526331, + "grad_norm": 0.12556962668895721, + "learning_rate": 9.945603292937311e-08, + "loss": 0.0098, + "step": 115820 + }, + { + "epoch": 33.700901949374455, + "grad_norm": 0.09136494994163513, + "learning_rate": 9.8981372801224e-08, + "loss": 0.0096, + "step": 115830 + }, + { + "epoch": 33.7038114634856, + "grad_norm": 0.14384965598583221, + "learning_rate": 9.850784430710135e-08, + "loss": 0.0103, + "step": 115840 + }, + { + "epoch": 33.70672097759674, + "grad_norm": 0.1305461972951889, + "learning_rate": 9.803544748296644e-08, + "loss": 0.0116, + "step": 115850 + }, + { + "epoch": 33.709630491707884, + "grad_norm": 0.13518047332763672, + "learning_rate": 9.756418236469555e-08, + "loss": 0.0091, + "step": 115860 + }, + { + "epoch": 33.71254000581903, + "grad_norm": 0.11526558548212051, + "learning_rate": 9.709404898807673e-08, + "loss": 0.0123, + "step": 115870 + }, + { + "epoch": 33.715449519930175, + "grad_norm": 0.07993466407060623, + "learning_rate": 9.662504738881473e-08, + "loss": 0.0107, + "step": 115880 + }, + { + "epoch": 33.71835903404131, + "grad_norm": 0.09202495217323303, + "learning_rate": 9.615717760252607e-08, + "loss": 0.0123, + "step": 115890 + }, + { + "epoch": 33.72126854815246, + "grad_norm": 0.08900069445371628, + "learning_rate": 9.569043966474233e-08, + "loss": 0.0115, + "step": 115900 + }, + { + "epoch": 33.724178062263604, + "grad_norm": 0.13634629547595978, + "learning_rate": 9.522483361091183e-08, + "loss": 0.0098, + "step": 115910 + }, + { + "epoch": 33.72708757637474, + "grad_norm": 0.1077101081609726, + "learning_rate": 9.476035947639128e-08, + "loss": 0.0103, + "step": 115920 + }, + { + "epoch": 33.72999709048589, + "grad_norm": 0.12857982516288757, + "learning_rate": 9.429701729645413e-08, + "loss": 0.0083, + "step": 115930 + }, + { + "epoch": 33.732906604597034, + "grad_norm": 0.10815288871526718, + "learning_rate": 9.383480710629056e-08, + "loss": 0.0087, + "step": 115940 + }, + { + "epoch": 33.73581611870818, + "grad_norm": 0.0890195369720459, + "learning_rate": 9.337372894100248e-08, + "loss": 0.0096, + "step": 115950 + }, + { + "epoch": 33.73872563281932, + "grad_norm": 0.0802723690867424, + "learning_rate": 9.291378283560358e-08, + "loss": 0.0097, + "step": 115960 + }, + { + "epoch": 33.74163514693046, + "grad_norm": 0.1255294382572174, + "learning_rate": 9.245496882502258e-08, + "loss": 0.0102, + "step": 115970 + }, + { + "epoch": 33.74454466104161, + "grad_norm": 0.10137420892715454, + "learning_rate": 9.199728694410825e-08, + "loss": 0.0115, + "step": 115980 + }, + { + "epoch": 33.74745417515275, + "grad_norm": 0.11305143684148788, + "learning_rate": 9.154073722761613e-08, + "loss": 0.0114, + "step": 115990 + }, + { + "epoch": 33.75036368926389, + "grad_norm": 0.09899803251028061, + "learning_rate": 9.108531971021516e-08, + "loss": 0.0092, + "step": 116000 + }, + { + "epoch": 33.75327320337504, + "grad_norm": 0.1314404457807541, + "learning_rate": 9.063103442649434e-08, + "loss": 0.0115, + "step": 116010 + }, + { + "epoch": 33.75618271748618, + "grad_norm": 0.1122356727719307, + "learning_rate": 9.01778814109544e-08, + "loss": 0.0104, + "step": 116020 + }, + { + "epoch": 33.75909223159732, + "grad_norm": 0.11307726800441742, + "learning_rate": 8.972586069800948e-08, + "loss": 0.0089, + "step": 116030 + }, + { + "epoch": 33.76200174570847, + "grad_norm": 0.148495152592659, + "learning_rate": 8.927497232198379e-08, + "loss": 0.0109, + "step": 116040 + }, + { + "epoch": 33.76491125981961, + "grad_norm": 0.12027471512556076, + "learning_rate": 8.88252163171216e-08, + "loss": 0.0098, + "step": 116050 + }, + { + "epoch": 33.76782077393075, + "grad_norm": 0.2736227810382843, + "learning_rate": 8.837659271758058e-08, + "loss": 0.0098, + "step": 116060 + }, + { + "epoch": 33.7707302880419, + "grad_norm": 0.11919072270393372, + "learning_rate": 8.79291015574285e-08, + "loss": 0.0112, + "step": 116070 + }, + { + "epoch": 33.77363980215304, + "grad_norm": 0.09010079503059387, + "learning_rate": 8.748274287064817e-08, + "loss": 0.0116, + "step": 116080 + }, + { + "epoch": 33.77654931626418, + "grad_norm": 0.1027393788099289, + "learning_rate": 8.703751669114246e-08, + "loss": 0.0094, + "step": 116090 + }, + { + "epoch": 33.779458830375326, + "grad_norm": 0.10634832084178925, + "learning_rate": 8.659342305271933e-08, + "loss": 0.0102, + "step": 116100 + }, + { + "epoch": 33.78236834448647, + "grad_norm": 0.11235425621271133, + "learning_rate": 8.615046198910514e-08, + "loss": 0.0102, + "step": 116110 + }, + { + "epoch": 33.78527785859762, + "grad_norm": 0.09590958803892136, + "learning_rate": 8.570863353394132e-08, + "loss": 0.0097, + "step": 116120 + }, + { + "epoch": 33.788187372708755, + "grad_norm": 0.10426904261112213, + "learning_rate": 8.526793772078101e-08, + "loss": 0.0099, + "step": 116130 + }, + { + "epoch": 33.7910968868199, + "grad_norm": 0.11052341759204865, + "learning_rate": 8.482837458309079e-08, + "loss": 0.0108, + "step": 116140 + }, + { + "epoch": 33.794006400931046, + "grad_norm": 0.10797492414712906, + "learning_rate": 8.43899441542556e-08, + "loss": 0.0108, + "step": 116150 + }, + { + "epoch": 33.796915915042185, + "grad_norm": 0.12319188565015793, + "learning_rate": 8.395264646756884e-08, + "loss": 0.0101, + "step": 116160 + }, + { + "epoch": 33.79982542915333, + "grad_norm": 0.11790390312671661, + "learning_rate": 8.351648155624225e-08, + "loss": 0.0113, + "step": 116170 + }, + { + "epoch": 33.802734943264475, + "grad_norm": 0.10497536510229111, + "learning_rate": 8.308144945339768e-08, + "loss": 0.0114, + "step": 116180 + }, + { + "epoch": 33.80564445737562, + "grad_norm": 0.10254865884780884, + "learning_rate": 8.26475501920737e-08, + "loss": 0.0114, + "step": 116190 + }, + { + "epoch": 33.80855397148676, + "grad_norm": 0.11398756504058838, + "learning_rate": 8.22147838052223e-08, + "loss": 0.0111, + "step": 116200 + }, + { + "epoch": 33.811463485597905, + "grad_norm": 0.060443777590990067, + "learning_rate": 8.178315032571049e-08, + "loss": 0.0119, + "step": 116210 + }, + { + "epoch": 33.81437299970905, + "grad_norm": 0.13265180587768555, + "learning_rate": 8.135264978631707e-08, + "loss": 0.0105, + "step": 116220 + }, + { + "epoch": 33.81728251382019, + "grad_norm": 0.12997151911258698, + "learning_rate": 8.092328221973256e-08, + "loss": 0.0108, + "step": 116230 + }, + { + "epoch": 33.820192027931334, + "grad_norm": 0.10820750147104263, + "learning_rate": 8.04950476585692e-08, + "loss": 0.0089, + "step": 116240 + }, + { + "epoch": 33.82310154204248, + "grad_norm": 0.08593837916851044, + "learning_rate": 8.006794613534762e-08, + "loss": 0.0091, + "step": 116250 + }, + { + "epoch": 33.826011056153625, + "grad_norm": 0.1591130793094635, + "learning_rate": 7.964197768250192e-08, + "loss": 0.009, + "step": 116260 + }, + { + "epoch": 33.82892057026476, + "grad_norm": 0.07031402736902237, + "learning_rate": 7.92171423323812e-08, + "loss": 0.0087, + "step": 116270 + }, + { + "epoch": 33.83183008437591, + "grad_norm": 0.07973747700452805, + "learning_rate": 7.879344011725132e-08, + "loss": 0.0102, + "step": 116280 + }, + { + "epoch": 33.834739598487054, + "grad_norm": 0.07641232013702393, + "learning_rate": 7.83708710692882e-08, + "loss": 0.0085, + "step": 116290 + }, + { + "epoch": 33.83764911259819, + "grad_norm": 0.12264532595872879, + "learning_rate": 7.794943522058285e-08, + "loss": 0.0095, + "step": 116300 + }, + { + "epoch": 33.84055862670934, + "grad_norm": 0.09446091949939728, + "learning_rate": 7.752913260313965e-08, + "loss": 0.0097, + "step": 116310 + }, + { + "epoch": 33.843468140820484, + "grad_norm": 0.06678557395935059, + "learning_rate": 7.710996324888142e-08, + "loss": 0.0094, + "step": 116320 + }, + { + "epoch": 33.84637765493163, + "grad_norm": 0.07549712806940079, + "learning_rate": 7.669192718963769e-08, + "loss": 0.0096, + "step": 116330 + }, + { + "epoch": 33.84928716904277, + "grad_norm": 0.14557233452796936, + "learning_rate": 7.62750244571564e-08, + "loss": 0.0106, + "step": 116340 + }, + { + "epoch": 33.85219668315391, + "grad_norm": 0.10042840242385864, + "learning_rate": 7.585925508309887e-08, + "loss": 0.0106, + "step": 116350 + }, + { + "epoch": 33.85510619726506, + "grad_norm": 0.15194033086299896, + "learning_rate": 7.544461909904154e-08, + "loss": 0.0101, + "step": 116360 + }, + { + "epoch": 33.8580157113762, + "grad_norm": 0.09395325183868408, + "learning_rate": 7.503111653647087e-08, + "loss": 0.0098, + "step": 116370 + }, + { + "epoch": 33.86092522548734, + "grad_norm": 0.1583215594291687, + "learning_rate": 7.46187474267901e-08, + "loss": 0.0117, + "step": 116380 + }, + { + "epoch": 33.86383473959849, + "grad_norm": 0.06531597673892975, + "learning_rate": 7.420751180131913e-08, + "loss": 0.0103, + "step": 116390 + }, + { + "epoch": 33.86674425370963, + "grad_norm": 0.0869482159614563, + "learning_rate": 7.379740969128468e-08, + "loss": 0.0089, + "step": 116400 + }, + { + "epoch": 33.86965376782077, + "grad_norm": 0.14112251996994019, + "learning_rate": 7.338844112783182e-08, + "loss": 0.0105, + "step": 116410 + }, + { + "epoch": 33.87256328193192, + "grad_norm": 0.0909552350640297, + "learning_rate": 7.298060614201907e-08, + "loss": 0.0091, + "step": 116420 + }, + { + "epoch": 33.87547279604306, + "grad_norm": 0.07424934953451157, + "learning_rate": 7.257390476482162e-08, + "loss": 0.0105, + "step": 116430 + }, + { + "epoch": 33.8783823101542, + "grad_norm": 0.1659509688615799, + "learning_rate": 7.216833702712145e-08, + "loss": 0.0118, + "step": 116440 + }, + { + "epoch": 33.88129182426535, + "grad_norm": 0.14968298375606537, + "learning_rate": 7.176390295972223e-08, + "loss": 0.0113, + "step": 116450 + }, + { + "epoch": 33.88420133837649, + "grad_norm": 0.1086306944489479, + "learning_rate": 7.136060259333776e-08, + "loss": 0.0102, + "step": 116460 + }, + { + "epoch": 33.88711085248764, + "grad_norm": 0.08870009332895279, + "learning_rate": 7.095843595859352e-08, + "loss": 0.0119, + "step": 116470 + }, + { + "epoch": 33.890020366598776, + "grad_norm": 0.11762186139822006, + "learning_rate": 7.055740308603342e-08, + "loss": 0.0109, + "step": 116480 + }, + { + "epoch": 33.89292988070992, + "grad_norm": 0.14728033542633057, + "learning_rate": 7.015750400611309e-08, + "loss": 0.0091, + "step": 116490 + }, + { + "epoch": 33.89583939482107, + "grad_norm": 0.11433687806129456, + "learning_rate": 6.975873874920158e-08, + "loss": 0.0124, + "step": 116500 + }, + { + "epoch": 33.898748908932205, + "grad_norm": 0.09055769443511963, + "learning_rate": 6.936110734558298e-08, + "loss": 0.0084, + "step": 116510 + }, + { + "epoch": 33.90165842304335, + "grad_norm": 0.11175937950611115, + "learning_rate": 6.896460982545483e-08, + "loss": 0.0102, + "step": 116520 + }, + { + "epoch": 33.904567937154496, + "grad_norm": 0.06953081488609314, + "learning_rate": 6.856924621892802e-08, + "loss": 0.0109, + "step": 116530 + }, + { + "epoch": 33.90747745126564, + "grad_norm": 0.1447017937898636, + "learning_rate": 6.817501655602687e-08, + "loss": 0.01, + "step": 116540 + }, + { + "epoch": 33.91038696537678, + "grad_norm": 0.11325273662805557, + "learning_rate": 6.77819208666941e-08, + "loss": 0.0093, + "step": 116550 + }, + { + "epoch": 33.913296479487926, + "grad_norm": 0.10696079581975937, + "learning_rate": 6.738995918077916e-08, + "loss": 0.0104, + "step": 116560 + }, + { + "epoch": 33.91620599359907, + "grad_norm": 0.1093638688325882, + "learning_rate": 6.699913152804992e-08, + "loss": 0.0125, + "step": 116570 + }, + { + "epoch": 33.91911550771021, + "grad_norm": 0.13859888911247253, + "learning_rate": 6.660943793818764e-08, + "loss": 0.009, + "step": 116580 + }, + { + "epoch": 33.922025021821355, + "grad_norm": 0.11925400048494339, + "learning_rate": 6.622087844078862e-08, + "loss": 0.0106, + "step": 116590 + }, + { + "epoch": 33.9249345359325, + "grad_norm": 0.08783354610204697, + "learning_rate": 6.58334530653576e-08, + "loss": 0.0102, + "step": 116600 + }, + { + "epoch": 33.927844050043646, + "grad_norm": 0.07420484721660614, + "learning_rate": 6.544716184131772e-08, + "loss": 0.0073, + "step": 116610 + }, + { + "epoch": 33.930753564154784, + "grad_norm": 0.17166242003440857, + "learning_rate": 6.506200479800883e-08, + "loss": 0.0115, + "step": 116620 + }, + { + "epoch": 33.93366307826593, + "grad_norm": 0.08142068982124329, + "learning_rate": 6.467798196467755e-08, + "loss": 0.0095, + "step": 116630 + }, + { + "epoch": 33.936572592377075, + "grad_norm": 0.10998876392841339, + "learning_rate": 6.429509337048889e-08, + "loss": 0.0091, + "step": 116640 + }, + { + "epoch": 33.939482106488214, + "grad_norm": 0.07880889624357224, + "learning_rate": 6.391333904452124e-08, + "loss": 0.0091, + "step": 116650 + }, + { + "epoch": 33.94239162059936, + "grad_norm": 0.12332505732774734, + "learning_rate": 6.353271901576641e-08, + "loss": 0.0103, + "step": 116660 + }, + { + "epoch": 33.945301134710505, + "grad_norm": 0.14866307377815247, + "learning_rate": 6.315323331312794e-08, + "loss": 0.0109, + "step": 116670 + }, + { + "epoch": 33.94821064882165, + "grad_norm": 0.07773975282907486, + "learning_rate": 6.27748819654278e-08, + "loss": 0.0086, + "step": 116680 + }, + { + "epoch": 33.95112016293279, + "grad_norm": 0.08883264660835266, + "learning_rate": 6.239766500139799e-08, + "loss": 0.0112, + "step": 116690 + }, + { + "epoch": 33.954029677043934, + "grad_norm": 0.09286633878946304, + "learning_rate": 6.202158244968559e-08, + "loss": 0.0131, + "step": 116700 + }, + { + "epoch": 33.95693919115508, + "grad_norm": 0.09783020615577698, + "learning_rate": 6.164663433885276e-08, + "loss": 0.0111, + "step": 116710 + }, + { + "epoch": 33.95984870526622, + "grad_norm": 0.10943752527236938, + "learning_rate": 6.127282069737172e-08, + "loss": 0.0094, + "step": 116720 + }, + { + "epoch": 33.96275821937736, + "grad_norm": 0.1339055895805359, + "learning_rate": 6.090014155363477e-08, + "loss": 0.0104, + "step": 116730 + }, + { + "epoch": 33.96566773348851, + "grad_norm": 0.07814466208219528, + "learning_rate": 6.05285969359426e-08, + "loss": 0.0092, + "step": 116740 + }, + { + "epoch": 33.968577247599654, + "grad_norm": 0.09179828315973282, + "learning_rate": 6.015818687251096e-08, + "loss": 0.011, + "step": 116750 + }, + { + "epoch": 33.97148676171079, + "grad_norm": 0.07684362679719925, + "learning_rate": 5.978891139147069e-08, + "loss": 0.0089, + "step": 116760 + }, + { + "epoch": 33.97439627582194, + "grad_norm": 0.07612022012472153, + "learning_rate": 5.942077052086603e-08, + "loss": 0.0096, + "step": 116770 + }, + { + "epoch": 33.977305789933084, + "grad_norm": 0.08214996755123138, + "learning_rate": 5.905376428865461e-08, + "loss": 0.0085, + "step": 116780 + }, + { + "epoch": 33.98021530404422, + "grad_norm": 0.100008524954319, + "learning_rate": 5.868789272270747e-08, + "loss": 0.0101, + "step": 116790 + }, + { + "epoch": 33.98312481815537, + "grad_norm": 0.10959991067647934, + "learning_rate": 5.832315585081072e-08, + "loss": 0.0107, + "step": 116800 + }, + { + "epoch": 33.98603433226651, + "grad_norm": 0.10395703464746475, + "learning_rate": 5.7959553700665526e-08, + "loss": 0.0106, + "step": 116810 + }, + { + "epoch": 33.98894384637765, + "grad_norm": 0.07858702540397644, + "learning_rate": 5.759708629988148e-08, + "loss": 0.0092, + "step": 116820 + }, + { + "epoch": 33.9918533604888, + "grad_norm": 0.11881201714277267, + "learning_rate": 5.723575367598821e-08, + "loss": 0.009, + "step": 116830 + }, + { + "epoch": 33.99476287459994, + "grad_norm": 0.09513601660728455, + "learning_rate": 5.687555585642546e-08, + "loss": 0.0092, + "step": 116840 + }, + { + "epoch": 33.99767238871109, + "grad_norm": 0.11122529953718185, + "learning_rate": 5.651649286854799e-08, + "loss": 0.011, + "step": 116850 + }, + { + "epoch": 34.000581902822226, + "grad_norm": 0.10640662163496017, + "learning_rate": 5.615856473962566e-08, + "loss": 0.0102, + "step": 116860 + }, + { + "epoch": 34.00349141693337, + "grad_norm": 0.17757506668567657, + "learning_rate": 5.580177149683841e-08, + "loss": 0.0108, + "step": 116870 + }, + { + "epoch": 34.00640093104452, + "grad_norm": 0.1049092561006546, + "learning_rate": 5.544611316728454e-08, + "loss": 0.0101, + "step": 116880 + }, + { + "epoch": 34.009310445155656, + "grad_norm": 0.09369147568941116, + "learning_rate": 5.5091589777972464e-08, + "loss": 0.0109, + "step": 116890 + }, + { + "epoch": 34.0122199592668, + "grad_norm": 0.11882789433002472, + "learning_rate": 5.4738201355825635e-08, + "loss": 0.0098, + "step": 116900 + }, + { + "epoch": 34.01512947337795, + "grad_norm": 0.06346320360898972, + "learning_rate": 5.438594792768259e-08, + "loss": 0.0095, + "step": 116910 + }, + { + "epoch": 34.01803898748909, + "grad_norm": 0.08617907762527466, + "learning_rate": 5.403482952029526e-08, + "loss": 0.0122, + "step": 116920 + }, + { + "epoch": 34.02094850160023, + "grad_norm": 0.08736756443977356, + "learning_rate": 5.3684846160327314e-08, + "loss": 0.0088, + "step": 116930 + }, + { + "epoch": 34.023858015711376, + "grad_norm": 0.09749122709035873, + "learning_rate": 5.33359978743575e-08, + "loss": 0.0098, + "step": 116940 + }, + { + "epoch": 34.02676752982252, + "grad_norm": 0.10177727788686752, + "learning_rate": 5.2988284688879616e-08, + "loss": 0.0109, + "step": 116950 + }, + { + "epoch": 34.02967704393366, + "grad_norm": 0.10103092342615128, + "learning_rate": 5.2641706630299214e-08, + "loss": 0.0093, + "step": 116960 + }, + { + "epoch": 34.032586558044805, + "grad_norm": 0.1248183324933052, + "learning_rate": 5.229626372493856e-08, + "loss": 0.0096, + "step": 116970 + }, + { + "epoch": 34.03549607215595, + "grad_norm": 0.08026779443025589, + "learning_rate": 5.1951955999028354e-08, + "loss": 0.0106, + "step": 116980 + }, + { + "epoch": 34.038405586267096, + "grad_norm": 0.12426438182592392, + "learning_rate": 5.1608783478721e-08, + "loss": 0.0104, + "step": 116990 + }, + { + "epoch": 34.041315100378235, + "grad_norm": 0.19637510180473328, + "learning_rate": 5.126674619007565e-08, + "loss": 0.0095, + "step": 117000 + }, + { + "epoch": 34.04422461448938, + "grad_norm": 0.08662225306034088, + "learning_rate": 5.092584415906654e-08, + "loss": 0.0102, + "step": 117010 + }, + { + "epoch": 34.047134128600526, + "grad_norm": 0.08472830802202225, + "learning_rate": 5.058607741158461e-08, + "loss": 0.0122, + "step": 117020 + }, + { + "epoch": 34.050043642711664, + "grad_norm": 0.08777248114347458, + "learning_rate": 5.0247445973432557e-08, + "loss": 0.0107, + "step": 117030 + }, + { + "epoch": 34.05295315682281, + "grad_norm": 0.10526551306247711, + "learning_rate": 4.990994987032649e-08, + "loss": 0.0121, + "step": 117040 + }, + { + "epoch": 34.055862670933955, + "grad_norm": 0.14186517894268036, + "learning_rate": 4.9573589127899225e-08, + "loss": 0.0102, + "step": 117050 + }, + { + "epoch": 34.0587721850451, + "grad_norm": 0.11162682622671127, + "learning_rate": 4.9238363771691994e-08, + "loss": 0.011, + "step": 117060 + }, + { + "epoch": 34.06168169915624, + "grad_norm": 0.09783017635345459, + "learning_rate": 4.890427382716611e-08, + "loss": 0.0085, + "step": 117070 + }, + { + "epoch": 34.064591213267384, + "grad_norm": 0.11255001276731491, + "learning_rate": 4.8571319319691276e-08, + "loss": 0.0121, + "step": 117080 + }, + { + "epoch": 34.06750072737853, + "grad_norm": 0.09026402235031128, + "learning_rate": 4.8239500274553924e-08, + "loss": 0.0083, + "step": 117090 + }, + { + "epoch": 34.07041024148967, + "grad_norm": 0.14565962553024292, + "learning_rate": 4.790881671695391e-08, + "loss": 0.0091, + "step": 117100 + }, + { + "epoch": 34.073319755600814, + "grad_norm": 0.11680352687835693, + "learning_rate": 4.75792686720028e-08, + "loss": 0.0094, + "step": 117110 + }, + { + "epoch": 34.07622926971196, + "grad_norm": 0.09266991913318634, + "learning_rate": 4.725085616473057e-08, + "loss": 0.011, + "step": 117120 + }, + { + "epoch": 34.079138783823105, + "grad_norm": 0.075086310505867, + "learning_rate": 4.6923579220075616e-08, + "loss": 0.0104, + "step": 117130 + }, + { + "epoch": 34.08204829793424, + "grad_norm": 0.16648785769939423, + "learning_rate": 4.659743786289306e-08, + "loss": 0.0131, + "step": 117140 + }, + { + "epoch": 34.08495781204539, + "grad_norm": 0.087431401014328, + "learning_rate": 4.627243211794974e-08, + "loss": 0.0105, + "step": 117150 + }, + { + "epoch": 34.087867326156534, + "grad_norm": 0.08936674892902374, + "learning_rate": 4.5948562009930915e-08, + "loss": 0.011, + "step": 117160 + }, + { + "epoch": 34.09077684026767, + "grad_norm": 0.11593475937843323, + "learning_rate": 4.562582756343026e-08, + "loss": 0.0087, + "step": 117170 + }, + { + "epoch": 34.09368635437882, + "grad_norm": 0.12605254352092743, + "learning_rate": 4.530422880295815e-08, + "loss": 0.0107, + "step": 117180 + }, + { + "epoch": 34.09659586848996, + "grad_norm": 0.25290194153785706, + "learning_rate": 4.498376575293672e-08, + "loss": 0.0115, + "step": 117190 + }, + { + "epoch": 34.09950538260111, + "grad_norm": 0.10813874751329422, + "learning_rate": 4.466443843770485e-08, + "loss": 0.0101, + "step": 117200 + }, + { + "epoch": 34.10241489671225, + "grad_norm": 0.07814999669790268, + "learning_rate": 4.434624688150979e-08, + "loss": 0.009, + "step": 117210 + }, + { + "epoch": 34.10532441082339, + "grad_norm": 0.1357916295528412, + "learning_rate": 4.402919110852055e-08, + "loss": 0.0095, + "step": 117220 + }, + { + "epoch": 34.10823392493454, + "grad_norm": 0.12209151685237885, + "learning_rate": 4.371327114281454e-08, + "loss": 0.0108, + "step": 117230 + }, + { + "epoch": 34.11114343904568, + "grad_norm": 0.12207381427288055, + "learning_rate": 4.33984870083809e-08, + "loss": 0.0081, + "step": 117240 + }, + { + "epoch": 34.11405295315682, + "grad_norm": 0.10789269208908081, + "learning_rate": 4.3084838729127165e-08, + "loss": 0.0097, + "step": 117250 + }, + { + "epoch": 34.11696246726797, + "grad_norm": 0.11414363980293274, + "learning_rate": 4.2772326328872624e-08, + "loss": 0.0093, + "step": 117260 + }, + { + "epoch": 34.11987198137911, + "grad_norm": 0.06184956803917885, + "learning_rate": 4.246094983135163e-08, + "loss": 0.0108, + "step": 117270 + }, + { + "epoch": 34.12278149549025, + "grad_norm": 0.09838224202394485, + "learning_rate": 4.215070926021025e-08, + "loss": 0.0105, + "step": 117280 + }, + { + "epoch": 34.1256910096014, + "grad_norm": 0.08186540752649307, + "learning_rate": 4.1841604639009654e-08, + "loss": 0.0094, + "step": 117290 + }, + { + "epoch": 34.12860052371254, + "grad_norm": 0.0737772285938263, + "learning_rate": 4.153363599122273e-08, + "loss": 0.0127, + "step": 117300 + }, + { + "epoch": 34.13151003782368, + "grad_norm": 0.1706942468881607, + "learning_rate": 4.122680334024076e-08, + "loss": 0.0128, + "step": 117310 + }, + { + "epoch": 34.134419551934826, + "grad_norm": 0.10951732099056244, + "learning_rate": 4.092110670936178e-08, + "loss": 0.0111, + "step": 117320 + }, + { + "epoch": 34.13732906604597, + "grad_norm": 0.11824709922075272, + "learning_rate": 4.0616546121803874e-08, + "loss": 0.0077, + "step": 117330 + }, + { + "epoch": 34.14023858015712, + "grad_norm": 0.10081804543733597, + "learning_rate": 4.031312160069689e-08, + "loss": 0.0119, + "step": 117340 + }, + { + "epoch": 34.143148094268255, + "grad_norm": 0.09655119478702545, + "learning_rate": 4.0010833169082384e-08, + "loss": 0.0094, + "step": 117350 + }, + { + "epoch": 34.1460576083794, + "grad_norm": 0.16364642977714539, + "learning_rate": 3.9709680849916995e-08, + "loss": 0.011, + "step": 117360 + }, + { + "epoch": 34.148967122490546, + "grad_norm": 0.11534536629915237, + "learning_rate": 3.94096646660741e-08, + "loss": 0.0094, + "step": 117370 + }, + { + "epoch": 34.151876636601685, + "grad_norm": 0.08805130422115326, + "learning_rate": 3.9110784640333796e-08, + "loss": 0.0101, + "step": 117380 + }, + { + "epoch": 34.15478615071283, + "grad_norm": 0.11790336668491364, + "learning_rate": 3.881304079539627e-08, + "loss": 0.0107, + "step": 117390 + }, + { + "epoch": 34.157695664823976, + "grad_norm": 0.12371298670768738, + "learning_rate": 3.851643315387343e-08, + "loss": 0.009, + "step": 117400 + }, + { + "epoch": 34.16060517893512, + "grad_norm": 0.08400320261716843, + "learning_rate": 3.822096173829059e-08, + "loss": 0.0092, + "step": 117410 + }, + { + "epoch": 34.16351469304626, + "grad_norm": 0.09926670044660568, + "learning_rate": 3.7926626571086474e-08, + "loss": 0.0104, + "step": 117420 + }, + { + "epoch": 34.166424207157405, + "grad_norm": 0.10164133459329605, + "learning_rate": 3.763342767461486e-08, + "loss": 0.0104, + "step": 117430 + }, + { + "epoch": 34.16933372126855, + "grad_norm": 0.15864846110343933, + "learning_rate": 3.734136507113961e-08, + "loss": 0.0099, + "step": 117440 + }, + { + "epoch": 34.17224323537969, + "grad_norm": 0.11817216128110886, + "learning_rate": 3.705043878284464e-08, + "loss": 0.0108, + "step": 117450 + }, + { + "epoch": 34.175152749490834, + "grad_norm": 0.09451248496770859, + "learning_rate": 3.676064883182062e-08, + "loss": 0.0091, + "step": 117460 + }, + { + "epoch": 34.17806226360198, + "grad_norm": 0.13426122069358826, + "learning_rate": 3.6471995240076605e-08, + "loss": 0.009, + "step": 117470 + }, + { + "epoch": 34.180971777713125, + "grad_norm": 0.1258108764886856, + "learning_rate": 3.6184478029533394e-08, + "loss": 0.0096, + "step": 117480 + }, + { + "epoch": 34.183881291824264, + "grad_norm": 0.09399426728487015, + "learning_rate": 3.589809722202853e-08, + "loss": 0.0071, + "step": 117490 + }, + { + "epoch": 34.18679080593541, + "grad_norm": 0.09082839637994766, + "learning_rate": 3.561285283930626e-08, + "loss": 0.0105, + "step": 117500 + }, + { + "epoch": 34.189700320046555, + "grad_norm": 0.1567094326019287, + "learning_rate": 3.5328744903032614e-08, + "loss": 0.0104, + "step": 117510 + }, + { + "epoch": 34.19260983415769, + "grad_norm": 0.10562194138765335, + "learning_rate": 3.504577343478366e-08, + "loss": 0.0109, + "step": 117520 + }, + { + "epoch": 34.19551934826884, + "grad_norm": 0.11113633215427399, + "learning_rate": 3.47639384560472e-08, + "loss": 0.0124, + "step": 117530 + }, + { + "epoch": 34.198428862379984, + "grad_norm": 0.12058890610933304, + "learning_rate": 3.448323998822611e-08, + "loss": 0.0095, + "step": 117540 + }, + { + "epoch": 34.20133837649112, + "grad_norm": 0.06785599887371063, + "learning_rate": 3.420367805264002e-08, + "loss": 0.0103, + "step": 117550 + }, + { + "epoch": 34.20424789060227, + "grad_norm": 0.09618471562862396, + "learning_rate": 3.392525267052027e-08, + "loss": 0.0092, + "step": 117560 + }, + { + "epoch": 34.20715740471341, + "grad_norm": 0.1349690705537796, + "learning_rate": 3.3647963863009943e-08, + "loss": 0.0121, + "step": 117570 + }, + { + "epoch": 34.21006691882456, + "grad_norm": 0.12566038966178894, + "learning_rate": 3.3371811651165543e-08, + "loss": 0.0114, + "step": 117580 + }, + { + "epoch": 34.2129764329357, + "grad_norm": 0.10836420953273773, + "learning_rate": 3.309679605596194e-08, + "loss": 0.0106, + "step": 117590 + }, + { + "epoch": 34.21588594704684, + "grad_norm": 0.12892164289951324, + "learning_rate": 3.282291709828411e-08, + "loss": 0.0104, + "step": 117600 + }, + { + "epoch": 34.21879546115799, + "grad_norm": 0.13935622572898865, + "learning_rate": 3.255017479893041e-08, + "loss": 0.0093, + "step": 117610 + }, + { + "epoch": 34.22170497526913, + "grad_norm": 0.09380965679883957, + "learning_rate": 3.2278569178614273e-08, + "loss": 0.0099, + "step": 117620 + }, + { + "epoch": 34.22461448938027, + "grad_norm": 0.09493963420391083, + "learning_rate": 3.2008100257962524e-08, + "loss": 0.0123, + "step": 117630 + }, + { + "epoch": 34.22752400349142, + "grad_norm": 0.08989895135164261, + "learning_rate": 3.1738768057517073e-08, + "loss": 0.0096, + "step": 117640 + }, + { + "epoch": 34.23043351760256, + "grad_norm": 0.11424672603607178, + "learning_rate": 3.147057259772823e-08, + "loss": 0.011, + "step": 117650 + }, + { + "epoch": 34.2333430317137, + "grad_norm": 0.1675957590341568, + "learning_rate": 3.120351389896636e-08, + "loss": 0.0105, + "step": 117660 + }, + { + "epoch": 34.23625254582485, + "grad_norm": 0.13017885386943817, + "learning_rate": 3.0937591981511916e-08, + "loss": 0.0117, + "step": 117670 + }, + { + "epoch": 34.23916205993599, + "grad_norm": 0.09103302657604218, + "learning_rate": 3.06728068655604e-08, + "loss": 0.0117, + "step": 117680 + }, + { + "epoch": 34.24207157404713, + "grad_norm": 0.0829547718167305, + "learning_rate": 3.0409158571219063e-08, + "loss": 0.0092, + "step": 117690 + }, + { + "epoch": 34.244981088158276, + "grad_norm": 0.09096657484769821, + "learning_rate": 3.0146647118513556e-08, + "loss": 0.01, + "step": 117700 + }, + { + "epoch": 34.24789060226942, + "grad_norm": 0.08313652873039246, + "learning_rate": 2.988527252737627e-08, + "loss": 0.0099, + "step": 117710 + }, + { + "epoch": 34.25080011638057, + "grad_norm": 0.1180175319314003, + "learning_rate": 2.9625034817657993e-08, + "loss": 0.0109, + "step": 117720 + }, + { + "epoch": 34.253709630491706, + "grad_norm": 0.093678317964077, + "learning_rate": 2.936593400912457e-08, + "loss": 0.0095, + "step": 117730 + }, + { + "epoch": 34.25661914460285, + "grad_norm": 0.13519267737865448, + "learning_rate": 2.9107970121448612e-08, + "loss": 0.0107, + "step": 117740 + }, + { + "epoch": 34.259528658714, + "grad_norm": 0.09447869658470154, + "learning_rate": 2.8851143174224437e-08, + "loss": 0.0082, + "step": 117750 + }, + { + "epoch": 34.262438172825135, + "grad_norm": 0.07271480560302734, + "learning_rate": 2.8595453186954777e-08, + "loss": 0.0101, + "step": 117760 + }, + { + "epoch": 34.26534768693628, + "grad_norm": 0.1576179713010788, + "learning_rate": 2.8340900179057438e-08, + "loss": 0.0118, + "step": 117770 + }, + { + "epoch": 34.268257201047426, + "grad_norm": 0.10761675983667374, + "learning_rate": 2.808748416986362e-08, + "loss": 0.0104, + "step": 117780 + }, + { + "epoch": 34.27116671515857, + "grad_norm": 0.06409718841314316, + "learning_rate": 2.783520517862126e-08, + "loss": 0.0104, + "step": 117790 + }, + { + "epoch": 34.27407622926971, + "grad_norm": 0.08117791265249252, + "learning_rate": 2.7584063224485034e-08, + "loss": 0.0103, + "step": 117800 + }, + { + "epoch": 34.276985743380855, + "grad_norm": 0.11177060008049011, + "learning_rate": 2.7334058326531354e-08, + "loss": 0.0104, + "step": 117810 + }, + { + "epoch": 34.279895257492, + "grad_norm": 0.12478521466255188, + "learning_rate": 2.7085190503745027e-08, + "loss": 0.0089, + "step": 117820 + }, + { + "epoch": 34.28280477160314, + "grad_norm": 0.145394429564476, + "learning_rate": 2.683745977502594e-08, + "loss": 0.0087, + "step": 117830 + }, + { + "epoch": 34.285714285714285, + "grad_norm": 0.07797399163246155, + "learning_rate": 2.659086615918571e-08, + "loss": 0.0088, + "step": 117840 + }, + { + "epoch": 34.28862379982543, + "grad_norm": 0.09328898042440414, + "learning_rate": 2.6345409674956024e-08, + "loss": 0.0105, + "step": 117850 + }, + { + "epoch": 34.291533313936576, + "grad_norm": 0.11104235798120499, + "learning_rate": 2.610109034097363e-08, + "loss": 0.0108, + "step": 117860 + }, + { + "epoch": 34.294442828047714, + "grad_norm": 0.11560434103012085, + "learning_rate": 2.5857908175793697e-08, + "loss": 0.0115, + "step": 117870 + }, + { + "epoch": 34.29735234215886, + "grad_norm": 0.13893350958824158, + "learning_rate": 2.5615863197884782e-08, + "loss": 0.0096, + "step": 117880 + }, + { + "epoch": 34.300261856270005, + "grad_norm": 0.1430227905511856, + "learning_rate": 2.537495542563051e-08, + "loss": 0.0118, + "step": 117890 + }, + { + "epoch": 34.30317137038114, + "grad_norm": 0.13175611197948456, + "learning_rate": 2.5135184877322914e-08, + "loss": 0.0102, + "step": 117900 + }, + { + "epoch": 34.30608088449229, + "grad_norm": 0.12965530157089233, + "learning_rate": 2.4896551571172432e-08, + "loss": 0.0117, + "step": 117910 + }, + { + "epoch": 34.308990398603434, + "grad_norm": 0.06797997653484344, + "learning_rate": 2.4659055525302898e-08, + "loss": 0.0092, + "step": 117920 + }, + { + "epoch": 34.31189991271458, + "grad_norm": 0.09895024448633194, + "learning_rate": 2.442269675774822e-08, + "loss": 0.0101, + "step": 117930 + }, + { + "epoch": 34.31480942682572, + "grad_norm": 0.07761017978191376, + "learning_rate": 2.4187475286460702e-08, + "loss": 0.0101, + "step": 117940 + }, + { + "epoch": 34.317718940936864, + "grad_norm": 0.1276521235704422, + "learning_rate": 2.395339112930106e-08, + "loss": 0.0084, + "step": 117950 + }, + { + "epoch": 34.32062845504801, + "grad_norm": 0.08563243597745895, + "learning_rate": 2.372044430405007e-08, + "loss": 0.0102, + "step": 117960 + }, + { + "epoch": 34.32353796915915, + "grad_norm": 0.10490110516548157, + "learning_rate": 2.348863482839525e-08, + "loss": 0.0113, + "step": 117970 + }, + { + "epoch": 34.32644748327029, + "grad_norm": 0.1698862910270691, + "learning_rate": 2.3257962719942516e-08, + "loss": 0.0088, + "step": 117980 + }, + { + "epoch": 34.32935699738144, + "grad_norm": 0.15196150541305542, + "learning_rate": 2.302842799620952e-08, + "loss": 0.0134, + "step": 117990 + }, + { + "epoch": 34.332266511492584, + "grad_norm": 0.1656128317117691, + "learning_rate": 2.2800030674627327e-08, + "loss": 0.0093, + "step": 118000 + }, + { + "epoch": 34.33517602560372, + "grad_norm": 0.06592484563589096, + "learning_rate": 2.2572770772543718e-08, + "loss": 0.0102, + "step": 118010 + }, + { + "epoch": 34.33808553971487, + "grad_norm": 0.1315012276172638, + "learning_rate": 2.234664830721489e-08, + "loss": 0.011, + "step": 118020 + }, + { + "epoch": 34.34099505382601, + "grad_norm": 0.09701064974069595, + "learning_rate": 2.2121663295813776e-08, + "loss": 0.0106, + "step": 118030 + }, + { + "epoch": 34.34390456793715, + "grad_norm": 0.058127790689468384, + "learning_rate": 2.189781575542671e-08, + "loss": 0.0086, + "step": 118040 + }, + { + "epoch": 34.3468140820483, + "grad_norm": 0.08422420173883438, + "learning_rate": 2.167510570305342e-08, + "loss": 0.0119, + "step": 118050 + }, + { + "epoch": 34.34972359615944, + "grad_norm": 0.13378971815109253, + "learning_rate": 2.145353315560872e-08, + "loss": 0.0117, + "step": 118060 + }, + { + "epoch": 34.35263311027059, + "grad_norm": 0.1317129135131836, + "learning_rate": 2.123309812991747e-08, + "loss": 0.0104, + "step": 118070 + }, + { + "epoch": 34.35554262438173, + "grad_norm": 0.10610689967870712, + "learning_rate": 2.1013800642722957e-08, + "loss": 0.0099, + "step": 118080 + }, + { + "epoch": 34.35845213849287, + "grad_norm": 0.10956978052854538, + "learning_rate": 2.079564071067519e-08, + "loss": 0.0098, + "step": 118090 + }, + { + "epoch": 34.36136165260402, + "grad_norm": 0.12094477564096451, + "learning_rate": 2.0578618350345912e-08, + "loss": 0.0103, + "step": 118100 + }, + { + "epoch": 34.364271166715156, + "grad_norm": 0.07987771183252335, + "learning_rate": 2.036273357821361e-08, + "loss": 0.0083, + "step": 118110 + }, + { + "epoch": 34.3671806808263, + "grad_norm": 0.10668838769197464, + "learning_rate": 2.0147986410675168e-08, + "loss": 0.0111, + "step": 118120 + }, + { + "epoch": 34.37009019493745, + "grad_norm": 0.08355999737977982, + "learning_rate": 1.9934376864039204e-08, + "loss": 0.0107, + "step": 118130 + }, + { + "epoch": 34.37299970904859, + "grad_norm": 0.12821514904499054, + "learning_rate": 1.9721904954527747e-08, + "loss": 0.0095, + "step": 118140 + }, + { + "epoch": 34.37590922315973, + "grad_norm": 0.09075549244880676, + "learning_rate": 1.951057069827622e-08, + "loss": 0.0095, + "step": 118150 + }, + { + "epoch": 34.378818737270876, + "grad_norm": 0.08650843054056168, + "learning_rate": 1.9300374111333452e-08, + "loss": 0.0091, + "step": 118160 + }, + { + "epoch": 34.38172825138202, + "grad_norm": 0.15410354733467102, + "learning_rate": 1.909131520966334e-08, + "loss": 0.011, + "step": 118170 + }, + { + "epoch": 34.38463776549316, + "grad_norm": 0.11082956194877625, + "learning_rate": 1.8883394009143186e-08, + "loss": 0.0104, + "step": 118180 + }, + { + "epoch": 34.387547279604306, + "grad_norm": 0.16346053779125214, + "learning_rate": 1.867661052556202e-08, + "loss": 0.0112, + "step": 118190 + }, + { + "epoch": 34.39045679371545, + "grad_norm": 0.12810876965522766, + "learning_rate": 1.8470964774623954e-08, + "loss": 0.0136, + "step": 118200 + }, + { + "epoch": 34.39336630782659, + "grad_norm": 0.09576006978750229, + "learning_rate": 1.8266456771948158e-08, + "loss": 0.0115, + "step": 118210 + }, + { + "epoch": 34.396275821937735, + "grad_norm": 0.06292126327753067, + "learning_rate": 1.806308653306221e-08, + "loss": 0.0104, + "step": 118220 + }, + { + "epoch": 34.39918533604888, + "grad_norm": 0.17538714408874512, + "learning_rate": 1.786085407341376e-08, + "loss": 0.0097, + "step": 118230 + }, + { + "epoch": 34.402094850160026, + "grad_norm": 0.10045831650495529, + "learning_rate": 1.7659759408358845e-08, + "loss": 0.01, + "step": 118240 + }, + { + "epoch": 34.405004364271164, + "grad_norm": 0.11893518269062042, + "learning_rate": 1.745980255317192e-08, + "loss": 0.0116, + "step": 118250 + }, + { + "epoch": 34.40791387838231, + "grad_norm": 0.1293673813343048, + "learning_rate": 1.7260983523035846e-08, + "loss": 0.0116, + "step": 118260 + }, + { + "epoch": 34.410823392493455, + "grad_norm": 0.09910473227500916, + "learning_rate": 1.70633023330502e-08, + "loss": 0.0104, + "step": 118270 + }, + { + "epoch": 34.413732906604594, + "grad_norm": 0.11084427684545517, + "learning_rate": 1.686675899822965e-08, + "loss": 0.0112, + "step": 118280 + }, + { + "epoch": 34.41664242071574, + "grad_norm": 0.0747825875878334, + "learning_rate": 1.6671353533497247e-08, + "loss": 0.0092, + "step": 118290 + }, + { + "epoch": 34.419551934826885, + "grad_norm": 0.06408549845218658, + "learning_rate": 1.647708595369446e-08, + "loss": 0.0109, + "step": 118300 + }, + { + "epoch": 34.42246144893803, + "grad_norm": 0.07297014445066452, + "learning_rate": 1.6283956273574484e-08, + "loss": 0.0105, + "step": 118310 + }, + { + "epoch": 34.42537096304917, + "grad_norm": 0.1530429869890213, + "learning_rate": 1.6091964507803926e-08, + "loss": 0.0097, + "step": 118320 + }, + { + "epoch": 34.428280477160314, + "grad_norm": 0.10838034003973007, + "learning_rate": 1.5901110670962783e-08, + "loss": 0.011, + "step": 118330 + }, + { + "epoch": 34.43118999127146, + "grad_norm": 0.0926162376999855, + "learning_rate": 1.5711394777546127e-08, + "loss": 0.0077, + "step": 118340 + }, + { + "epoch": 34.4340995053826, + "grad_norm": 0.08453333377838135, + "learning_rate": 1.552281684196244e-08, + "loss": 0.0108, + "step": 118350 + }, + { + "epoch": 34.43700901949374, + "grad_norm": 0.10396793484687805, + "learning_rate": 1.5335376878531927e-08, + "loss": 0.0114, + "step": 118360 + }, + { + "epoch": 34.43991853360489, + "grad_norm": 0.07732348144054413, + "learning_rate": 1.5149074901488202e-08, + "loss": 0.0092, + "step": 118370 + }, + { + "epoch": 34.442828047716034, + "grad_norm": 0.12018019706010818, + "learning_rate": 1.4963910924981616e-08, + "loss": 0.0107, + "step": 118380 + }, + { + "epoch": 34.44573756182717, + "grad_norm": 0.10822589695453644, + "learning_rate": 1.4779884963072587e-08, + "loss": 0.0103, + "step": 118390 + }, + { + "epoch": 34.44864707593832, + "grad_norm": 0.10365793853998184, + "learning_rate": 1.4596997029736603e-08, + "loss": 0.0095, + "step": 118400 + }, + { + "epoch": 34.451556590049464, + "grad_norm": 0.07404147833585739, + "learning_rate": 1.4415247138865883e-08, + "loss": 0.0125, + "step": 118410 + }, + { + "epoch": 34.4544661041606, + "grad_norm": 0.14231297373771667, + "learning_rate": 1.423463530425939e-08, + "loss": 0.0086, + "step": 118420 + }, + { + "epoch": 34.45737561827175, + "grad_norm": 0.10408905148506165, + "learning_rate": 1.4055161539636152e-08, + "loss": 0.0094, + "step": 118430 + }, + { + "epoch": 34.46028513238289, + "grad_norm": 0.11197537183761597, + "learning_rate": 1.3876825858623598e-08, + "loss": 0.0106, + "step": 118440 + }, + { + "epoch": 34.46319464649404, + "grad_norm": 0.13348592817783356, + "learning_rate": 1.3699628274767561e-08, + "loss": 0.0131, + "step": 118450 + }, + { + "epoch": 34.46610416060518, + "grad_norm": 0.08384132385253906, + "learning_rate": 1.352356880152228e-08, + "loss": 0.0097, + "step": 118460 + }, + { + "epoch": 34.46901367471632, + "grad_norm": 0.10205044597387314, + "learning_rate": 1.3348647452260387e-08, + "loss": 0.0101, + "step": 118470 + }, + { + "epoch": 34.47192318882747, + "grad_norm": 0.12525640428066254, + "learning_rate": 1.317486424026626e-08, + "loss": 0.0089, + "step": 118480 + }, + { + "epoch": 34.474832702938606, + "grad_norm": 0.09847959876060486, + "learning_rate": 1.3002219178736007e-08, + "loss": 0.0123, + "step": 118490 + }, + { + "epoch": 34.47774221704975, + "grad_norm": 0.08540048450231552, + "learning_rate": 1.283071228078081e-08, + "loss": 0.0099, + "step": 118500 + }, + { + "epoch": 34.4806517311609, + "grad_norm": 0.10365810245275497, + "learning_rate": 1.2660343559426913e-08, + "loss": 0.0094, + "step": 118510 + }, + { + "epoch": 34.48356124527204, + "grad_norm": 0.12010686099529266, + "learning_rate": 1.24911130276123e-08, + "loss": 0.0099, + "step": 118520 + }, + { + "epoch": 34.48647075938318, + "grad_norm": 0.20123255252838135, + "learning_rate": 1.2323020698190023e-08, + "loss": 0.0112, + "step": 118530 + }, + { + "epoch": 34.489380273494326, + "grad_norm": 0.10049042105674744, + "learning_rate": 1.2156066583923208e-08, + "loss": 0.0088, + "step": 118540 + }, + { + "epoch": 34.49228978760547, + "grad_norm": 0.2056250423192978, + "learning_rate": 1.1990250697490047e-08, + "loss": 0.011, + "step": 118550 + }, + { + "epoch": 34.49519930171661, + "grad_norm": 0.12318383157253265, + "learning_rate": 1.1825573051487127e-08, + "loss": 0.0094, + "step": 118560 + }, + { + "epoch": 34.498108815827756, + "grad_norm": 0.12447409331798553, + "learning_rate": 1.1662033658419447e-08, + "loss": 0.0107, + "step": 118570 + }, + { + "epoch": 34.5010183299389, + "grad_norm": 0.09476412087678909, + "learning_rate": 1.149963253070374e-08, + "loss": 0.0101, + "step": 118580 + }, + { + "epoch": 34.50392784405005, + "grad_norm": 0.10836581140756607, + "learning_rate": 1.1338369680676809e-08, + "loss": 0.0113, + "step": 118590 + }, + { + "epoch": 34.506837358161185, + "grad_norm": 0.10033656656742096, + "learning_rate": 1.1178245120583851e-08, + "loss": 0.0115, + "step": 118600 + }, + { + "epoch": 34.50974687227233, + "grad_norm": 0.09494254738092422, + "learning_rate": 1.1019258862585146e-08, + "loss": 0.0099, + "step": 118610 + }, + { + "epoch": 34.512656386383476, + "grad_norm": 0.07287470996379852, + "learning_rate": 1.0861410918756032e-08, + "loss": 0.0109, + "step": 118620 + }, + { + "epoch": 34.515565900494614, + "grad_norm": 0.11417483538389206, + "learning_rate": 1.070470130108192e-08, + "loss": 0.0095, + "step": 118630 + }, + { + "epoch": 34.51847541460576, + "grad_norm": 0.07777088135480881, + "learning_rate": 1.0549130021464959e-08, + "loss": 0.0099, + "step": 118640 + }, + { + "epoch": 34.521384928716905, + "grad_norm": 0.07565118372440338, + "learning_rate": 1.0394697091720695e-08, + "loss": 0.0114, + "step": 118650 + }, + { + "epoch": 34.52429444282805, + "grad_norm": 0.07155228406190872, + "learning_rate": 1.0241402523576416e-08, + "loss": 0.011, + "step": 118660 + }, + { + "epoch": 34.52720395693919, + "grad_norm": 0.055621493607759476, + "learning_rate": 1.0089246328672807e-08, + "loss": 0.0084, + "step": 118670 + }, + { + "epoch": 34.530113471050335, + "grad_norm": 0.07400696724653244, + "learning_rate": 9.938228518567294e-09, + "loss": 0.0089, + "step": 118680 + }, + { + "epoch": 34.53302298516148, + "grad_norm": 0.09509781002998352, + "learning_rate": 9.788349104725703e-09, + "loss": 0.0099, + "step": 118690 + }, + { + "epoch": 34.53593249927262, + "grad_norm": 0.097788006067276, + "learning_rate": 9.639608098533925e-09, + "loss": 0.0082, + "step": 118700 + }, + { + "epoch": 34.538842013383764, + "grad_norm": 0.20659074187278748, + "learning_rate": 9.49200551128626e-09, + "loss": 0.0121, + "step": 118710 + }, + { + "epoch": 34.54175152749491, + "grad_norm": 0.3662491738796234, + "learning_rate": 9.345541354190413e-09, + "loss": 0.0101, + "step": 118720 + }, + { + "epoch": 34.544661041606055, + "grad_norm": 0.11423538625240326, + "learning_rate": 9.200215638372477e-09, + "loss": 0.0114, + "step": 118730 + }, + { + "epoch": 34.54757055571719, + "grad_norm": 0.11399926245212555, + "learning_rate": 9.056028374866965e-09, + "loss": 0.0094, + "step": 118740 + }, + { + "epoch": 34.55048006982834, + "grad_norm": 0.08401496708393097, + "learning_rate": 8.912979574625113e-09, + "loss": 0.0089, + "step": 118750 + }, + { + "epoch": 34.553389583939484, + "grad_norm": 0.13674820959568024, + "learning_rate": 8.771069248508234e-09, + "loss": 0.0101, + "step": 118760 + }, + { + "epoch": 34.55629909805062, + "grad_norm": 0.08315252512693405, + "learning_rate": 8.630297407297704e-09, + "loss": 0.0091, + "step": 118770 + }, + { + "epoch": 34.55920861216177, + "grad_norm": 0.0846276804804802, + "learning_rate": 8.490664061679977e-09, + "loss": 0.0102, + "step": 118780 + }, + { + "epoch": 34.562118126272914, + "grad_norm": 0.1452612727880478, + "learning_rate": 8.352169222259897e-09, + "loss": 0.0095, + "step": 118790 + }, + { + "epoch": 34.56502764038406, + "grad_norm": 0.11144470423460007, + "learning_rate": 8.214812899559054e-09, + "loss": 0.01, + "step": 118800 + }, + { + "epoch": 34.5679371544952, + "grad_norm": 0.07877112925052643, + "learning_rate": 8.078595104004106e-09, + "loss": 0.0105, + "step": 118810 + }, + { + "epoch": 34.57084666860634, + "grad_norm": 0.0921756848692894, + "learning_rate": 7.943515845941773e-09, + "loss": 0.0107, + "step": 118820 + }, + { + "epoch": 34.57375618271749, + "grad_norm": 0.11763215065002441, + "learning_rate": 7.80957513563052e-09, + "loss": 0.0107, + "step": 118830 + }, + { + "epoch": 34.57666569682863, + "grad_norm": 0.15027830004692078, + "learning_rate": 7.676772983243874e-09, + "loss": 0.0115, + "step": 118840 + }, + { + "epoch": 34.57957521093977, + "grad_norm": 0.06850502640008926, + "learning_rate": 7.545109398863769e-09, + "loss": 0.0089, + "step": 118850 + }, + { + "epoch": 34.58248472505092, + "grad_norm": 0.10216091573238373, + "learning_rate": 7.414584392490542e-09, + "loss": 0.0085, + "step": 118860 + }, + { + "epoch": 34.585394239162056, + "grad_norm": 0.10604722797870636, + "learning_rate": 7.285197974039592e-09, + "loss": 0.0101, + "step": 118870 + }, + { + "epoch": 34.5883037532732, + "grad_norm": 0.1198640838265419, + "learning_rate": 7.156950153333064e-09, + "loss": 0.0107, + "step": 118880 + }, + { + "epoch": 34.59121326738435, + "grad_norm": 0.11040479689836502, + "learning_rate": 7.029840940111498e-09, + "loss": 0.0091, + "step": 118890 + }, + { + "epoch": 34.59412278149549, + "grad_norm": 0.13447979092597961, + "learning_rate": 6.90387034402884e-09, + "loss": 0.0092, + "step": 118900 + }, + { + "epoch": 34.59703229560663, + "grad_norm": 0.10018313676118851, + "learning_rate": 6.779038374652435e-09, + "loss": 0.0114, + "step": 118910 + }, + { + "epoch": 34.59994180971778, + "grad_norm": 0.10307541489601135, + "learning_rate": 6.655345041459704e-09, + "loss": 0.0095, + "step": 118920 + }, + { + "epoch": 34.60285132382892, + "grad_norm": 0.0813726857304573, + "learning_rate": 6.5327903538464625e-09, + "loss": 0.0103, + "step": 118930 + }, + { + "epoch": 34.60576083794006, + "grad_norm": 0.08839832246303558, + "learning_rate": 6.411374321120267e-09, + "loss": 0.0112, + "step": 118940 + }, + { + "epoch": 34.608670352051206, + "grad_norm": 0.1343439221382141, + "learning_rate": 6.2910969525020735e-09, + "loss": 0.0084, + "step": 118950 + }, + { + "epoch": 34.61157986616235, + "grad_norm": 0.1306726634502411, + "learning_rate": 6.1719582571229115e-09, + "loss": 0.0095, + "step": 118960 + }, + { + "epoch": 34.6144893802735, + "grad_norm": 0.08313843607902527, + "learning_rate": 6.0539582440338745e-09, + "loss": 0.0098, + "step": 118970 + }, + { + "epoch": 34.617398894384635, + "grad_norm": 0.10071597248315811, + "learning_rate": 5.937096922196128e-09, + "loss": 0.0117, + "step": 118980 + }, + { + "epoch": 34.62030840849578, + "grad_norm": 0.09036128968000412, + "learning_rate": 5.821374300482574e-09, + "loss": 0.0094, + "step": 118990 + }, + { + "epoch": 34.623217922606926, + "grad_norm": 0.11580657958984375, + "learning_rate": 5.706790387682848e-09, + "loss": 0.009, + "step": 119000 + }, + { + "epoch": 34.626127436718065, + "grad_norm": 0.11457178741693497, + "learning_rate": 5.593345192498322e-09, + "loss": 0.0108, + "step": 119010 + }, + { + "epoch": 34.62903695082921, + "grad_norm": 0.083550825715065, + "learning_rate": 5.481038723545439e-09, + "loss": 0.0091, + "step": 119020 + }, + { + "epoch": 34.631946464940356, + "grad_norm": 0.11705878376960754, + "learning_rate": 5.369870989352377e-09, + "loss": 0.0106, + "step": 119030 + }, + { + "epoch": 34.6348559790515, + "grad_norm": 0.0785648450255394, + "learning_rate": 5.259841998362381e-09, + "loss": 0.0091, + "step": 119040 + }, + { + "epoch": 34.63776549316264, + "grad_norm": 0.06956597417593002, + "learning_rate": 5.150951758928768e-09, + "loss": 0.0089, + "step": 119050 + }, + { + "epoch": 34.640675007273785, + "grad_norm": 0.12715494632720947, + "learning_rate": 5.043200279324922e-09, + "loss": 0.0102, + "step": 119060 + }, + { + "epoch": 34.64358452138493, + "grad_norm": 0.10495741665363312, + "learning_rate": 4.936587567730965e-09, + "loss": 0.0104, + "step": 119070 + }, + { + "epoch": 34.64649403549607, + "grad_norm": 0.10611774772405624, + "learning_rate": 4.831113632243755e-09, + "loss": 0.0098, + "step": 119080 + }, + { + "epoch": 34.649403549607214, + "grad_norm": 0.14300262928009033, + "learning_rate": 4.726778480875216e-09, + "loss": 0.0098, + "step": 119090 + }, + { + "epoch": 34.65231306371836, + "grad_norm": 0.14386862516403198, + "learning_rate": 4.6235821215473425e-09, + "loss": 0.0109, + "step": 119100 + }, + { + "epoch": 34.655222577829505, + "grad_norm": 0.1282382756471634, + "learning_rate": 4.521524562097201e-09, + "loss": 0.0095, + "step": 119110 + }, + { + "epoch": 34.658132091940644, + "grad_norm": 0.09122148901224136, + "learning_rate": 4.420605810275258e-09, + "loss": 0.0097, + "step": 119120 + }, + { + "epoch": 34.66104160605179, + "grad_norm": 0.08903498947620392, + "learning_rate": 4.320825873747048e-09, + "loss": 0.0115, + "step": 119130 + }, + { + "epoch": 34.663951120162935, + "grad_norm": 0.11634386330842972, + "learning_rate": 4.222184760088177e-09, + "loss": 0.0126, + "step": 119140 + }, + { + "epoch": 34.66686063427407, + "grad_norm": 0.09226131439208984, + "learning_rate": 4.124682476792652e-09, + "loss": 0.0105, + "step": 119150 + }, + { + "epoch": 34.66977014838522, + "grad_norm": 0.08647021651268005, + "learning_rate": 4.028319031261218e-09, + "loss": 0.0096, + "step": 119160 + }, + { + "epoch": 34.672679662496364, + "grad_norm": 0.08562306314706802, + "learning_rate": 3.933094430816353e-09, + "loss": 0.0093, + "step": 119170 + }, + { + "epoch": 34.67558917660751, + "grad_norm": 0.08379307389259338, + "learning_rate": 3.8390086826856076e-09, + "loss": 0.0101, + "step": 119180 + }, + { + "epoch": 34.67849869071865, + "grad_norm": 0.12314529716968536, + "learning_rate": 3.746061794016597e-09, + "loss": 0.0101, + "step": 119190 + }, + { + "epoch": 34.68140820482979, + "grad_norm": 0.09673997759819031, + "learning_rate": 3.654253771868676e-09, + "loss": 0.0108, + "step": 119200 + }, + { + "epoch": 34.68431771894094, + "grad_norm": 0.08222205191850662, + "learning_rate": 3.563584623211269e-09, + "loss": 0.0095, + "step": 119210 + }, + { + "epoch": 34.68722723305208, + "grad_norm": 0.10926046967506409, + "learning_rate": 3.474054354932199e-09, + "loss": 0.009, + "step": 119220 + }, + { + "epoch": 34.69013674716322, + "grad_norm": 0.07038076967000961, + "learning_rate": 3.3856629738310273e-09, + "loss": 0.0099, + "step": 119230 + }, + { + "epoch": 34.69304626127437, + "grad_norm": 0.09532737731933594, + "learning_rate": 3.298410486619052e-09, + "loss": 0.0106, + "step": 119240 + }, + { + "epoch": 34.695955775385514, + "grad_norm": 0.09032963216304779, + "learning_rate": 3.212296899924305e-09, + "loss": 0.0098, + "step": 119250 + }, + { + "epoch": 34.69886528949665, + "grad_norm": 0.17020151019096375, + "learning_rate": 3.127322220283224e-09, + "loss": 0.0114, + "step": 119260 + }, + { + "epoch": 34.7017748036078, + "grad_norm": 0.09876697510480881, + "learning_rate": 3.043486454152311e-09, + "loss": 0.0095, + "step": 119270 + }, + { + "epoch": 34.70468431771894, + "grad_norm": 0.10236819833517075, + "learning_rate": 2.96078960789814e-09, + "loss": 0.0086, + "step": 119280 + }, + { + "epoch": 34.70759383183008, + "grad_norm": 0.08062794059515, + "learning_rate": 2.879231687799022e-09, + "loss": 0.0114, + "step": 119290 + }, + { + "epoch": 34.71050334594123, + "grad_norm": 0.08825766295194626, + "learning_rate": 2.798812700050002e-09, + "loss": 0.0113, + "step": 119300 + }, + { + "epoch": 34.71341286005237, + "grad_norm": 0.0811181589961052, + "learning_rate": 2.7195326507595264e-09, + "loss": 0.0101, + "step": 119310 + }, + { + "epoch": 34.71632237416352, + "grad_norm": 0.10285723954439163, + "learning_rate": 2.6413915459461146e-09, + "loss": 0.0092, + "step": 119320 + }, + { + "epoch": 34.719231888274656, + "grad_norm": 0.0981115847826004, + "learning_rate": 2.5643893915450187e-09, + "loss": 0.0102, + "step": 119330 + }, + { + "epoch": 34.7221414023858, + "grad_norm": 0.11174961924552917, + "learning_rate": 2.4885261934032287e-09, + "loss": 0.0121, + "step": 119340 + }, + { + "epoch": 34.72505091649695, + "grad_norm": 0.10954218357801437, + "learning_rate": 2.413801957284467e-09, + "loss": 0.0085, + "step": 119350 + }, + { + "epoch": 34.727960430608086, + "grad_norm": 0.11883050203323364, + "learning_rate": 2.3402166888608635e-09, + "loss": 0.0093, + "step": 119360 + }, + { + "epoch": 34.73086994471923, + "grad_norm": 0.08560995012521744, + "learning_rate": 2.2677703937229453e-09, + "loss": 0.0107, + "step": 119370 + }, + { + "epoch": 34.73377945883038, + "grad_norm": 0.06578007340431213, + "learning_rate": 2.1964630773713136e-09, + "loss": 0.0127, + "step": 119380 + }, + { + "epoch": 34.73668897294152, + "grad_norm": 0.13476215302944183, + "learning_rate": 2.1262947452199697e-09, + "loss": 0.0101, + "step": 119390 + }, + { + "epoch": 34.73959848705266, + "grad_norm": 0.08878269791603088, + "learning_rate": 2.057265402601316e-09, + "loss": 0.011, + "step": 119400 + }, + { + "epoch": 34.742508001163806, + "grad_norm": 0.08242614567279816, + "learning_rate": 1.989375054754494e-09, + "loss": 0.0104, + "step": 119410 + }, + { + "epoch": 34.74541751527495, + "grad_norm": 0.1021231859922409, + "learning_rate": 1.92262370683538e-09, + "loss": 0.0106, + "step": 119420 + }, + { + "epoch": 34.74832702938609, + "grad_norm": 0.23770035803318024, + "learning_rate": 1.8570113639149177e-09, + "loss": 0.01, + "step": 119430 + }, + { + "epoch": 34.751236543497235, + "grad_norm": 0.1370684653520584, + "learning_rate": 1.7925380309757878e-09, + "loss": 0.0097, + "step": 119440 + }, + { + "epoch": 34.75414605760838, + "grad_norm": 0.1272241622209549, + "learning_rate": 1.7292037129140737e-09, + "loss": 0.0129, + "step": 119450 + }, + { + "epoch": 34.757055571719526, + "grad_norm": 0.08738737553358078, + "learning_rate": 1.6670084145392617e-09, + "loss": 0.0099, + "step": 119460 + }, + { + "epoch": 34.759965085830665, + "grad_norm": 0.08117683976888657, + "learning_rate": 1.6059521405742406e-09, + "loss": 0.0107, + "step": 119470 + }, + { + "epoch": 34.76287459994181, + "grad_norm": 0.14487554132938385, + "learning_rate": 1.546034895656967e-09, + "loss": 0.0104, + "step": 119480 + }, + { + "epoch": 34.765784114052956, + "grad_norm": 0.08285524696111679, + "learning_rate": 1.487256684337135e-09, + "loss": 0.009, + "step": 119490 + }, + { + "epoch": 34.768693628164094, + "grad_norm": 0.07557661086320877, + "learning_rate": 1.4296175110778409e-09, + "loss": 0.0094, + "step": 119500 + }, + { + "epoch": 34.77160314227524, + "grad_norm": 0.088335782289505, + "learning_rate": 1.3731173802572495e-09, + "loss": 0.0109, + "step": 119510 + }, + { + "epoch": 34.774512656386385, + "grad_norm": 0.13575346767902374, + "learning_rate": 1.3177562961669276e-09, + "loss": 0.0103, + "step": 119520 + }, + { + "epoch": 34.77742217049753, + "grad_norm": 0.13809922337532043, + "learning_rate": 1.2635342630085144e-09, + "loss": 0.0094, + "step": 119530 + }, + { + "epoch": 34.78033168460867, + "grad_norm": 0.06719715148210526, + "learning_rate": 1.210451284903713e-09, + "loss": 0.011, + "step": 119540 + }, + { + "epoch": 34.783241198719814, + "grad_norm": 0.13888758420944214, + "learning_rate": 1.1585073658809675e-09, + "loss": 0.0112, + "step": 119550 + }, + { + "epoch": 34.78615071283096, + "grad_norm": 0.1405944973230362, + "learning_rate": 1.1077025098854554e-09, + "loss": 0.0104, + "step": 119560 + }, + { + "epoch": 34.7890602269421, + "grad_norm": 0.09846333414316177, + "learning_rate": 1.058036720775757e-09, + "loss": 0.0122, + "step": 119570 + }, + { + "epoch": 34.791969741053244, + "grad_norm": 0.10489160567522049, + "learning_rate": 1.00951000232552e-09, + "loss": 0.0086, + "step": 119580 + }, + { + "epoch": 34.79487925516439, + "grad_norm": 0.08173870295286179, + "learning_rate": 9.621223582167993e-10, + "loss": 0.0103, + "step": 119590 + }, + { + "epoch": 34.79778876927553, + "grad_norm": 0.0826789066195488, + "learning_rate": 9.158737920500481e-10, + "loss": 0.0092, + "step": 119600 + }, + { + "epoch": 34.80069828338667, + "grad_norm": 0.10311046987771988, + "learning_rate": 8.707643073357918e-10, + "loss": 0.0112, + "step": 119610 + }, + { + "epoch": 34.80360779749782, + "grad_norm": 0.08247760683298111, + "learning_rate": 8.26793907502954e-10, + "loss": 0.0118, + "step": 119620 + }, + { + "epoch": 34.806517311608964, + "grad_norm": 0.08376117050647736, + "learning_rate": 7.839625958905306e-10, + "loss": 0.0101, + "step": 119630 + }, + { + "epoch": 34.8094268257201, + "grad_norm": 0.10634606331586838, + "learning_rate": 7.422703757475891e-10, + "loss": 0.0098, + "step": 119640 + }, + { + "epoch": 34.81233633983125, + "grad_norm": 0.05403178930282593, + "learning_rate": 7.017172502432612e-10, + "loss": 0.0098, + "step": 119650 + }, + { + "epoch": 34.81524585394239, + "grad_norm": 0.05598849803209305, + "learning_rate": 6.623032224584158e-10, + "loss": 0.0097, + "step": 119660 + }, + { + "epoch": 34.81815536805353, + "grad_norm": 0.12693478167057037, + "learning_rate": 6.240282953823284e-10, + "loss": 0.0121, + "step": 119670 + }, + { + "epoch": 34.82106488216468, + "grad_norm": 0.13713228702545166, + "learning_rate": 5.868924719243385e-10, + "loss": 0.0108, + "step": 119680 + }, + { + "epoch": 34.82397439627582, + "grad_norm": 0.06625296175479889, + "learning_rate": 5.508957549055227e-10, + "loss": 0.0087, + "step": 119690 + }, + { + "epoch": 34.82688391038697, + "grad_norm": 0.0971950963139534, + "learning_rate": 5.16038147058695e-10, + "loss": 0.0093, + "step": 119700 + }, + { + "epoch": 34.829793424498106, + "grad_norm": 0.07810936123132706, + "learning_rate": 4.82319651030072e-10, + "loss": 0.0097, + "step": 119710 + }, + { + "epoch": 34.83270293860925, + "grad_norm": 0.12055648118257523, + "learning_rate": 4.4974026938093824e-10, + "loss": 0.0115, + "step": 119720 + }, + { + "epoch": 34.8356124527204, + "grad_norm": 0.11349856108427048, + "learning_rate": 4.183000045859808e-10, + "loss": 0.0097, + "step": 119730 + }, + { + "epoch": 34.838521966831536, + "grad_norm": 0.16184769570827484, + "learning_rate": 3.8799885903162415e-10, + "loss": 0.0128, + "step": 119740 + }, + { + "epoch": 34.84143148094268, + "grad_norm": 0.09972194582223892, + "learning_rate": 3.5883683502102583e-10, + "loss": 0.011, + "step": 119750 + }, + { + "epoch": 34.84434099505383, + "grad_norm": 0.20967313647270203, + "learning_rate": 3.308139347674155e-10, + "loss": 0.0129, + "step": 119760 + }, + { + "epoch": 34.84725050916497, + "grad_norm": 0.0812348946928978, + "learning_rate": 3.039301603990907e-10, + "loss": 0.0104, + "step": 119770 + }, + { + "epoch": 34.85016002327611, + "grad_norm": 0.10218928009271622, + "learning_rate": 2.7818551395775163e-10, + "loss": 0.0079, + "step": 119780 + }, + { + "epoch": 34.853069537387256, + "grad_norm": 0.07040470093488693, + "learning_rate": 2.535799973985009e-10, + "loss": 0.0097, + "step": 119790 + }, + { + "epoch": 34.8559790514984, + "grad_norm": 0.1281152367591858, + "learning_rate": 2.3011361258984397e-10, + "loss": 0.0086, + "step": 119800 + }, + { + "epoch": 34.85888856560954, + "grad_norm": 0.08776123821735382, + "learning_rate": 2.077863613153541e-10, + "loss": 0.011, + "step": 119810 + }, + { + "epoch": 34.861798079720685, + "grad_norm": 0.11854979395866394, + "learning_rate": 1.8659824526867653e-10, + "loss": 0.0098, + "step": 119820 + }, + { + "epoch": 34.86470759383183, + "grad_norm": 0.06477639079093933, + "learning_rate": 1.6654926606018973e-10, + "loss": 0.0076, + "step": 119830 + }, + { + "epoch": 34.867617107942976, + "grad_norm": 0.1340726613998413, + "learning_rate": 1.4763942521200946e-10, + "loss": 0.0115, + "step": 119840 + }, + { + "epoch": 34.870526622054115, + "grad_norm": 0.11204170435667038, + "learning_rate": 1.2986872415965413e-10, + "loss": 0.0112, + "step": 119850 + }, + { + "epoch": 34.87343613616526, + "grad_norm": 0.10497820377349854, + "learning_rate": 1.1323716425371e-10, + "loss": 0.01, + "step": 119860 + }, + { + "epoch": 34.876345650276406, + "grad_norm": 0.08375784754753113, + "learning_rate": 9.77447467565007e-11, + "loss": 0.0088, + "step": 119870 + }, + { + "epoch": 34.879255164387544, + "grad_norm": 0.1111755520105362, + "learning_rate": 8.33914728454177e-11, + "loss": 0.0104, + "step": 119880 + }, + { + "epoch": 34.88216467849869, + "grad_norm": 0.082642562687397, + "learning_rate": 7.017734360958983e-11, + "loss": 0.0119, + "step": 119890 + }, + { + "epoch": 34.885074192609835, + "grad_norm": 0.0982419028878212, + "learning_rate": 5.810236005154845e-11, + "loss": 0.0089, + "step": 119900 + }, + { + "epoch": 34.88798370672098, + "grad_norm": 0.10877154767513275, + "learning_rate": 4.716652309222358e-11, + "loss": 0.0114, + "step": 119910 + }, + { + "epoch": 34.89089322083212, + "grad_norm": 0.06789495795965195, + "learning_rate": 3.7369833557621137e-11, + "loss": 0.0105, + "step": 119920 + }, + { + "epoch": 34.893802734943264, + "grad_norm": 0.08195552974939346, + "learning_rate": 2.8712292193810995e-11, + "loss": 0.012, + "step": 119930 + }, + { + "epoch": 34.89671224905441, + "grad_norm": 0.09670759737491608, + "learning_rate": 2.11938996586003e-11, + "loss": 0.0087, + "step": 119940 + }, + { + "epoch": 34.89962176316555, + "grad_norm": 0.10317841917276382, + "learning_rate": 1.4814656521533464e-11, + "loss": 0.01, + "step": 119950 + }, + { + "epoch": 34.902531277276694, + "grad_norm": 0.15020567178726196, + "learning_rate": 9.57456326888817e-12, + "loss": 0.0107, + "step": 119960 + }, + { + "epoch": 34.90544079138784, + "grad_norm": 0.1235247328877449, + "learning_rate": 5.4736202986793714e-12, + "loss": 0.009, + "step": 119970 + }, + { + "epoch": 34.908350305498985, + "grad_norm": 0.11598283797502518, + "learning_rate": 2.5118279189939585e-12, + "loss": 0.0098, + "step": 119980 + }, + { + "epoch": 34.91125981961012, + "grad_norm": 0.11540386825799942, + "learning_rate": 6.891863596480974e-13, + "loss": 0.0083, + "step": 119990 + }, + { + "epoch": 34.91416933372127, + "grad_norm": 0.09390421211719513, + "learning_rate": 5.69575553388546e-15, + "loss": 0.0119, + "step": 120000 + }, + { + "epoch": 34.91416933372127, + "step": 120000, + "total_flos": 0.0, + "train_loss": 0.020937973015631238, + "train_runtime": 224174.0781, + "train_samples_per_second": 68.518, + "train_steps_per_second": 0.535 + } + ], + "logging_steps": 10, + "max_steps": 120000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 20000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 64, + "trial_name": null, + "trial_params": null +} diff --git a/training_args.bin b/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..0865001f90ce8f1d2cc8c81d865be6d0e9aec684 --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59b878e9aa947d76003868019bbb9cc1e1990529da1e72a1628cd5c9469b7158 +size 5777